Recent science on how organizations work.
Peer-reviewed work from the world's top institutions, each with a plain-English takeaway. Via OpenAlex · updated Oct 13, 2026.
The briefFit is the new lever: agent pairings, AI roles, and skills that surface with tenure
The strongest experiment this cycle treats the human-AI pairing, not the model, as the unit of design. In a preregistered study, 1,258 people worked with AI agents prompted to exhibit different Big Five personalities; the teams produced 7,266 real display ads, and quality depended on the pairing itself, with extraverted humans matched to conscientious AI producing the worst output, a result carried into a field test on X with nearly 5 million impressions. Acceptance follows the same logic of specificity: across eleven European countries and 5,250 workers, acceptance of workplace AI proved strongly role-dependent, an assistant and a supervisor drawing different answers from the same workforce. An expert panel on accessible work adds the caution that the consequences of these systems are structural, organizational and emotional, not merely technical, able to remove barriers for disabled workers and erect new ones at once.
Hiring research converged on a related point: what is legible at the start of the employment deal is thin, and the real information surfaces later. Management Science evidence from job postings shows that writing culture into the ad works as a sorting device, raising worker inflows, interview-to-offer conversion and offer acceptance, especially where seekers cannot learn the culture elsewhere. Matched UK employee-employer data finds that social skills, initially opaque to both employee and employer, are revealed over tenure through complementarity with co-workers and drive wage progression for less educated workers. Both results sit naturally beside the wage model already on this radar, in which hiring pay barely echoes forward and the long-term deal is sustained at costly renegotiation moments.
The informal fabric of organizations supplied the cautionary notes. Two field surveys and an experiment show that where gossip about peers' performance is prevalent, employees inflate reported performance to protect their social image, so the grapevine can quietly corrupt the metrics it appears to police. A qualitative synthesis of six healthcare studies (854 participants) maps leadership's influence on psychological safety as multiple interconnected mechanisms rather than a single behavior, extending the best-backed theme on this radar. And a four-study comparison finds nonprofits retain employees better than for-profits when they install employee-centered practices (benefits, participation, training), a reminder that the same practice draws different loyalty depending on the organization that hosts it.
Read the full radar, 102 papers ↓Working papers and institutional reports land weeks before journals index them. This is the research of the week, before peer review catches up.
- 150-plus specialized AI agents under 8 "superagents": an architecture for governing multi-agent HR before it fragments
The Leader's Guide to Multi-Agent AI Architecture maps 150-plus AI agents attracting CHRO interest into 8 coordinating superagent families, with six deliberate governance decisions: orchestration, monitoring and security, data reasoning rules, human oversight, shared capability libraries, and build versus buy versus extend. Cases include DBS Bank, Mastercard, ASICS and Costa Coffee.
For consultants: A ready-made framing for the client with 40 pilots and no architecture: agent sprawl is an organizational governance problem for business leaders, not an IT configuration task. - Firms that exit collective bargaining gain only 0.3 to 0.6 percent more weekly hours: the restrictive-agreement story is mostly myth
German establishment data from 2008 to 2024 show firms leaving collective agreements raise weekly hours by just 0.3 to 0.6 percent (8 to 13 minutes) with no meaningful rise in weekend work. That contradicts employer-survey claims that working-time constraints are a main driver of exit from collective bargaining.
For consultants: A counterweight when a client blames labor agreements for inflexibility: revealed behavior says the binding constraint is rarely the negotiated schedule, so find the real bottleneck in work design before renegotiating contracts. - 535 million employment records: AI talent concentrates in leading firms, and network position predicts enterprise value beyond headcount
Mapping interfirm AI talent flows across roughly 535 million employment records in 58 countries (2010 to 2022), leading firms increasingly dominate AI talent acquisition, though positions in the network remain more fluid than in general labor flows. Centrality in the AI talent network is associated with higher enterprise value beyond what workforce size explains.
For consultants: Reframe the AI talent conversation from headcount to network position: where a client sits in the circulation of AI professionals, who they hire from and who they lose to, is itself a measurable competitive asset. - 80 percent say AI boosts their own productivity, 6 percent of firms are high performers: the organization is the bottleneck
The org-design follow-on to September's State of AI survey argues agentic AI relaxes the three constraints that shaped organizational design (human capacity, expertise, coordination), yet only 37 percent of firms see EBIT impact and coordinator roles have grown 1.5 to 2 times faster than line roles over five years. Nearly three quarters of AI high performers substantially redesigned workflows, against about one quarter of everyone else.
For consultants: The page to put in front of a client who bought copilots and stopped: the dividend only arrives when freed capacity gets an explicit destination, and jobs, teams and layers are redesigned around that choice. - Fixed wages cut productivity 8.7 percent, but attract slightly better workers
An experiment isolates the two forces pay design has to trade off: fixed wages lower worker productivity by about 8.74 percent versus piece rates through moral hazard, while attracting marginally higher-productivity workers through selection. The welfare math is close enough that a targeted tax on piece rates could raise revenue at a social cost as low as 0.88 dollars per dollar.
For consultants: The counterweight to the retention case for stable pay: read it next to the HBR finding that variable pay drives turnover, and the honest answer is that no comp lever is free. - Rater effects of 0.79 SD can swamp real AI-agent improvement of 0.53 SD: evaluation programs mislead without rater controls
Across 2,611 scored interviews with a deployed voice-and-video interviewing agent, differences in reviewer severity reached 0.79 standard deviations, enough to distort observed trends; after adjusting for raters, the agent genuinely improved 0.53 standard deviations from March to August 2026, corroborated by interview-related support tickets falling about 10 percent monthly.
For consultants: If a client judges agentic AI pilots with human review panels, the measurement system itself may be the biggest source of noise: insist on rater calibration before any go or no-go decision. - The AI buildout needs nearly 9 trillion dollars by 2032: one 1 GW AI campus alone costs 41 billion
A full 188 GW US datacenter buildout by 2032 would require almost 9 trillion dollars, about 3.2 percent of GDP over 2025 to 2032. As hyperscalers exhaust internal cash, financing shifts to leases, joint ventures, project debt, private credit and securitization, importing asset-level leverage and obsolescence risk into the AI stack.
For consultants: When a client's AI roadmap assumes cheap, abundant compute, show the financing side: build price and availability risk into the business case. - 29 percent of 2026 US STEM PhD dissertations contain AI-written text, and AI-heavy writers leave academia
AI-generated writing is absent in dissertations before 2023, appears in 29 percent of those filed in 2026, and is growing fast, with higher rates among international students and lower-ranked institutions. Graduates with AI-heavy dissertations are less likely to continue in academic research, and the authors argue AI may be displacing the very work that builds expertise.
For consultants: Hard evidence for the expertise-atrophy conversation: the next cohort of knowledge workers arrives with AI-assisted credentials, so hiring screens and early-career development need to test for underlying skill, not polished output. - New machinery raises demand for high-STEM workers 45 percent, but entirely at external suppliers, not the investing firm
Belgian data linking investment, supply chains and worker skills show the skill bias of automation lands wholly outside the investing firm: after machinery investments, purchases from high-STEM suppliers rise 45 percent, three times other skill groups, while in-house skilled hiring barely moves. Rising machinery per worker from 2003 to 2021 raised the high-STEM wage premium by 3.1 points, invisible in payroll-only data.
For consultants: When clients automate, the skills bill does not disappear, it moves into the supplier ecosystem: workforce planning should count the external STEM capacity a technology program will consume. - Companies risk losing exactly their most AI-savvy people
49,364 workers across 48 countries: 64 percent used AI at work this year and daily GenAI use jumped from 14 to 22 percent, yet access to learning fell from 59 to 51 percent. Among scarce-skill workers with strong AI capability, 29 percent say they are likely to change employer within a year.
For consultants: The retention argument for autonomy: the people most able to leave are the ones whose learning access is being cut. - When less is more: managing AI adoption with adaptive incentive design
A medical-device firm mandated 200 AI queries a month for 5,000 employees and got threshold-gaming, with a third of queries redundant. Cutting the target to 100 dropped volume 30 percent, but 90 percent of the decline was gaming, legitimate use held, and sales rose about 7 percent.
For consultants: Hard evidence that usage quotas buy compliance theater; lowering the mandate and trusting judgment raised real adoption and performance. - Working with agentic "teammates": a new actor collides with the human ecosystem of work
An in-situ study of a persistent, proactive AI agent deployed as a standing team member in a large tech company. It disrupted workflow protocols, relational boundaries and the allocation of trust, forcing teams to renegotiate norms.
For consultants: When an agent becomes a teammate rather than a tool, the team's social contract (who decides, who is trusted) has to be redesigned explicitly. - Improving today, narrowing tomorrow: collective learning, diversity and generativity
A formal model of firms learning from shared best-practice repertoires, generative AI included: copying leaders consumes diversity fast, while shared abstractions plus local judgment preserve future option space.
For consultants: The formal case for the Semco thesis that practices must be adapted through local judgment, never installed as templates. - Agents are scaling, earnings are not: the operating-model gap in one survey
40 percent of $1B-plus organizations are scaling AI agents in at least one function, up from 27 percent a year earlier, yet the share reporting AI contributes to EBIT is flat at 37 percent and high performers remain stuck near 6 percent.
For consultants: The single best data point for arguing that agent deployment without operating-model redesign produces motion, not money.
- Psychological safety & voiceLeadership and psychological safety in healthcare: a qualitative evidence synthesis
- Remote, hybrid & flexible workAI-coaching and well-being in sales: a theory of tools perspective
- AI, agents & the future of managementHow Artificial Intelligence Shapes Human Narratives
- Wellbeing, autonomy & motivationStress, resilience, and recovery: assessing a physical restorative space within a multimodal support program for hospital workers
- Credentials & skills-based hiringWage Adjustment in Efficient Long-Term Employment Relationships
- Remote, hybrid & flexible workCommuting matters: Association between work travel and perceived health and well-being in Canada
When Is It Fairer to Allocate Resources for Training to Lower Performing Versus Higher Performing Employees?
Journal of Accounting Research — The University of Texas at Austin · Emory University
Managers splitting scarce training budgets face a fairness trap that depends on context, not on the allocation itself. In an interactive experiment, employees judged the same decision by the prevailing norm: where egalitarian norms dominated, as when people had little control over their tasks, lower performers treated training as their due and reacted badly when it went to top performers, while meritocratic norms reversed the reaction. For consultants: before advising where development money goes, read which fairness norm the environment has installed, because task autonomy shifts the norm itself.
Can Nonprofits Gain an Advantage in Employee Retention Through Management Practices? A Comparative Lens for Managing Nonprofit Versus for‐Profit Employees
Personnel Psychology — Chinese University of Hong Kong · University of Florida
Nonprofits may hold a structural retention edge over for-profits, but only when they install employee-centered management practices: benefits, participation and training. Across four studies, including archival Statistics Canada survey data, nonprofit employees rated their organizations more positively than for-profit employees did when the organization implemented more of these practices, consistent with the theory that mission-driven settings convert them into firm-specific attachment. For consultants: participation and training are not generic retention levers; their pull depends on the kind of organization that hosts them, so calibrate the retention playbook to sector before copying it across clients.
Odd One Out: A Configural Perspective on Team Mental Models
Small Group Research — Erasmus University Rotterdam · Film Independent
Cognitive isolation, where one team member holds a task mental model dissimilar from everyone else's, undermined performance in a study of 79 student teams running a business simulation, but only when member epistemic motivation was low. Average alignment can mask a single outlier whose divergent understanding drags results. For consultants: when diagnosing a struggling self-managed team, look for the odd one out in how members understand the task, not just overall agreement.
Coordination and balanced communication support collective problem-solving in organic teams
npj Complexity — Corvinus University of Budapest · Brigham and Women's Hospital
Teams that solved escape-room problems most effectively coordinated throughout the task, kept communication balanced across members, and addressed tasks promptly as they became solvable, while members of less effective teams drifted into isolated work. The evidence comes from high-frequency observational data on naturally occurring teams rather than lab self-reports. For consultants: balanced airtime and fast pickup of newly unblocked work are observable markers you can use to assess and coach collective problem-solving in self-organizing teams.
What is the role of direct and indirect participation in organisational occupational health interventions? A configural perspective
Work & Stress — Umeå University · Karolinska Institutet
Direct and indirect employee participation turned out to depend on each other in organisational occupational health interventions across 64 Danish preschools, analysed with coincidence analysis to capture causal complexity that regression misses. Neither channel alone determines whether an intervention takes hold. For consultants: when designing workplace change programmes, combine hands-on employee involvement with representative structures instead of treating one as a substitute for the other.
Discipline and justify: Ideas and the making of economic hierarchy
Finance and Society — Copenhagen Business School
Contemporary economic hierarchy is sustained by ideas that justify it, from the political project of austerity after the First World War to newer neoliberal arguments that rationalize inequality through the language of science and biological variation. Two recent books reviewed here trace how capitalism has been conceptualized, defended, and sustained. For consultants: arguments defending steep hierarchy are historically constructed, and knowing their lineage helps you counter them when making the case for workplace democracy.
Alternative Organisations and (In)equality: Institutional Voids and Subjectivities in a Refugee Women’s Cooperative
Work Employment and Society — Usak University · University of Edinburgh
Inequality persisted inside a refugee women's cooperative in Turkey because formal institutional voids (no supportive legal framework, a top-down governance model) combined with internalised subjectivities shaped by patriarchy and authoritarianism to constrain members' expectations of voice. Alternative organising alone did not undo the hierarchies members brought with them. For consultants: cooperative structures need deliberate design for voice and supportive institutions, or they quietly reproduce the inequalities they were meant to replace.
The Dual Face of Workplace AI: A Typology of Digital Work Configurations and Job Quality Implications in Europe
New Technology Work and Employment — Institute for Employment Studies · University of Warwick
Workers who use AI tools intensively while keeping autonomy (Digitally Empowered) earn 21 percentage points more relative to country medians, while Heavily Controlled workers face reduced autonomy and work intensification without compensating wage premiums, across 50,462 workers in 27 EU countries. Job quality depends on the configuration of AI use and algorithmic control, not on AI exposure alone. For consultants: the question for clients is not whether to adopt AI but who controls it, since empowered use and algorithmic management produce opposite job quality outcomes.
Closing the Gap, Carrying the Burden: How Gender and Prosocial Motivation Shape Workplace Recognition
Personnel Psychology — The Ohio State University · Tsinghua University
Across a multi-source field study of 275 employees and two experiments, women were less likely than men to hold central positions in advice networks, and centrality is the peer-conferred signal that feeds leadership emergence and performance ratings. High prosocial motivation closed the gap by giving women a role-congruent path to becoming valued advice sources, which is parity carried as an extra burden. For consultants: recognition systems that read informal networks inherit their biases; look at who the advice actually flows through before calling a talent review objective.
The Fabricated Front: Generative AI and the Opacity of Workplace Performance
The Paris Journal on AI & Digital Ethics — University of California, Berkeley
Generative AI decouples observable output from human engagement at work, a condition of effort opacity that makes interactional cues less diagnostic and weakens the reciprocal exchange sustaining collaborative trust. Analysis of 1,250 interview transcripts identifies five mechanisms through which workplace fronts are reorganized. For consultants: as AI spreads, effort becomes invisible, so help clients shift recognition and trust toward verified outcomes and explicit norms about AI use.
The Process Model of Anxiety: A Framework and Review of Anxiety in the Workplace
Annual Review of Organizational Psychology and Organizational Behavior — University of Toronto · Virginia Tech
Workplace anxiety is best understood as a process with identifiable stages (appraisal, activation, processing, fatigue, behavior) rather than a static, uniformly harmful state, and it can support vigilance, preparation, and safety as well as undermine performance, wellbeing, and relationships. Feedback, regulation, recovery, and context modulate the whole sequence. For consultants: target the stage where anxiety goes wrong for a given team instead of treating all anxiety as a problem to eliminate.
The workplace as a panoptic machine: linguistic surveillance and migrants in Australia
International Journal of the Sociology of Language — The University of Melbourne
Migrant employees in Australia report that their use of languages other than English is hierarchically observed and morally judged by managers and co-workers, a linguistic surveillance that operates as a mechanism of workplace power. A pragmatic code of English-here positions multilingual workers as the linguistic Other. For consultants: language norms are a hidden hierarchy, and auditing how they are policed belongs in any serious inclusion or flattening effort.
The social-class gap in negotiation: Lower-class individuals negotiate less and face more backlash
Proceedings of the National Academy of Sciences — Hong Kong University of Science and Technology · Massachusetts Institute of Technology
Lower-class individuals are less likely to initiate negotiations than higher-class counterparts and face more backlash when they do, a behavioral pathway that compounds the class ceiling in organizations even when credentials are comparable, shown across four studies totaling 10,211 participants. Asking is not equally available to everyone. For consultants: pay systems that rely on employees negotiating quietly favor the already privileged; transparent, negotiation-free structures remove that hidden advantage.
Sexual harassment at work: Targets’ perspectives on prevention and response
PLoS ONE — University of Arkansas at Fayetteville · Stanford Medicine
Workers who experienced sexual harassment credit two organizational practices with preventing and answering it: setting clear expectations that harassment will not be tolerated, and responding effectively through an independent entity to whom victims can report incidents. The account is grounded in interviews with thirty US workers, most from the service industry. For consultants: an independent reporting channel and visibly enforced expectations are what targets themselves say works, a useful check on policy-heavy but response-light programmes.
Workplace surveillance as a form of discrimination
Alternative Law Journal — The University of Melbourne
Workplace surveillance, increasingly widespread and invasive in Australian workplaces, poses a significant risk of enabling or enacting discrimination, and discrimination law offers a route to challenge it. Its impacts fall unevenly across workers. For consultants: monitoring tools carry legal as well as cultural risk, and treating surveillance as a discrimination issue reframes the client conversation beyond privacy.
How does job autonomy influence employee helping behavior?
European Management Journal — University of Birmingham · University of Manchester
Job autonomy predicts employee helping behavior through organization-based self-esteem: autonomous employees value themselves more as organization members and help colleagues more, in two samples of service workers and their supervisors in China. The link weakens among employees high in neuroticism. For consultants: autonomy pays social dividends beyond individual performance, one more argument for loosening control when building collaborative cultures.
Leadership and psychological safety in healthcare: a qualitative evidence synthesis
Leadership in health services — Tan Tock Seng Hospital · National University of Singapore
A qualitative evidence synthesis pulls together six studies from the USA, UK and Ireland (854 participants: nurses, physicians, allied health professionals and administrators across hospitals, intensive care units, primary care and Veterans Health Administration facilities) and finds that leadership influences psychological safety through multiple interconnected mechanisms rather than any single behavior. For consultants: in safety-critical settings the leadership-to-safety link now has a cross-context qualitative map; treat it as a system of mutually reinforcing leader behaviors to be designed together, not a single trainable skill.
Implementing Psychological Safety in the Emergency Department Guided by the CFIR 2.0 Framework: Introducing the SAFE-ED Programme
Journal of Critical Care & Emergency Medicine — SingHealth · SingHealth Duke-NUS Academic Medical Centre
Psychological safety in emergency departments correlates directly with improved diagnostic accuracy, lower burnout rates, and lower patient mortality, yet it is not experienced uniformly across teams. The SAFE-ED programme translates it into a structured implementation pathway guided by the CFIR 2.0 framework for these fast-paced, multi-tiered hierarchies. For consultants: even in the most hierarchical, high-pressure settings, psychological safety can be implemented through a staged framework rather than culture talk alone.
Organization-Level Inclusion: Evidence of Positive Effects for Both LGBT and Non-LGBT Employees
Journal of Business Ethics — University of California, Berkeley · University of Utah
Organization-level inclusion initiatives closed the psychological safety gap between LGBT and non-LGBT employees and also produced positive effects for non-LGBT staff, across an archival study, a preregistered time-lagged study, and a randomized experiment. Inclusion signals at the organizational level shape the psychological climate for everyone. For consultants: inclusion efforts are not zero-sum, and evidence from three study designs supports pitching them as a whole-workforce climate investment.
Examining the experience of psychological safety in football academies: Players’ perspective
Sports Coaching Review — University of Edinburgh · Edinburgh Napier University
Fourteen academy footballers described psychological safety as embedded but rarely named: it lived in reciprocal voice, learning from mistakes and authority-figure behaviour, and it thinned out precisely during transitions, poor results and heightened emotional demand. Safety was conditional on the scoreboard. For consultants: the moments a client's culture is tested, new arrivals and bad quarters, are exactly when safety evaporates unless leaders behave deliberately.
Information processing, mutual understanding, and organization design in healthcare
Strategic Management Journal — University of California, Irvine · Irvine University
Electronic health record implementation in a large US hospital system increased decentralization of decisions, and the effect was amplified when care team members shared mutual understanding of each other, task processes, and the problem-solving context. Information access alone does not decentralize; shared interpretation does. For consultants: pushing decisions down requires building mutual understanding across the team, not just deploying systems that make information available.
Feedback in the Wild: A Pilot Study of Workplace Feedback Practices in Mechanical Engineering
Proceedings of the Canadian Engineering Education Association (CEEA) — University of British Columbia
Team evaluation in mechanical engineering workplaces is common and frequently embedded in project debriefs and informal discussions, while individual performance feedback is typically structured, multi-purpose, and draws on multiple sources, in preliminary data from 63 organizations. Formal and informal feedback operate at different levels of the same workplace. For consultants: map both the informal team debrief culture and the formal individual channels before redesigning a client's feedback system.
Talent Hoarding in Organizations
American Economic Review — LMU Klinikum · Ludwig-Maximilians-Universität München
Three-fourths of managers self-report hoarding talented workers, hoarding intensifies under performance-related pay, larger teams, and talent visibility, and exposure to it deters internal job applications and inhibits career progression. Managers evaluated on team performance have a built-in incentive to keep their best people hidden. For consultants: internal mobility problems are often incentive problems, so expect talent hoarding wherever managers are scored on team results and redesign the incentive first.
Burnout Prevention in Male Team Sport: Intervention Implications from Reciprocal Associations with Psychological Safety
Journal of Sport Psychology in Action — Keio University · The University of Tokyo
A three-wave study of 126 team athletes found the two arrows run at different speeds: psychological safety reduced emotional and interpersonal exhaustion over two months, while interpersonal exhaustion undermined safety within one. The authors propose building safety at team formation, monitoring exhaustion monthly and stepping in individually when it appears. For consultants: safety erodes faster than it builds, so the monitoring cadence matters as much as the intervention.
Leading through relationships: the project Manager's role in project-based organizations
Management Decision — University of Aberdeen · Loughborough University
Project managers in project-based organizations influence team performance and cohesion through the relationships they establish and maintain, examined through in-depth interviews with project managers and team members. Relational structures within teams, and how leaders handle interpersonal challenges, carry project outcomes. For consultants: in flat, project-based settings the real leadership work is relational maintenance, a competence worth assessing and developing explicitly.
A paradox of brokerage: alter distrust, network self-promotion, and the willingness to accept referrals from brokers
Organizational Behavior and Human Decision Processes — University College London · Bocconi University
Brokers positioned across structural holes are well placed to refer disconnected people to each other, yet those same networks undermine the trust recipients need before accepting a referral; network self-promotion partly resolves the paradox, substituting for reputational cues where shared contacts cannot supply them. Evidence spans a field study and two experiments. For consultants: a broad network can itself breed distrust, so advertise your matchmaking competence deliberately when brokering introductions between organizations.
Performance Manipulation as a Reaction to the Prevalence of Gossip: The Role of the Self‐Monitoring Trait
Contemporary Accounting Research — KU Leuven · École Supérieure des Sciences Économiques et Commerciales
Gossip is usually cast as informal discipline that deters misconduct, but two field surveys and an experiment find a second channel running the other way: where gossip about peers' performance is prevalent, employees fear negative social evaluation and inflate their reported performance to protect their social image. The effect is weaker for high self-monitors, who manage impressions through other means. For consultants: a gossip-heavy culture can quietly corrupt the performance data itself, which matters for any client running self-reported metrics or peer-visible dashboards.
Cultural conformity reverses the effectiveness of trust-based management: an Ising-inspired model of the Care-Dare dilemma
Frontiers in Physics — The University of Tokyo
Whether investing in trust before demanding performance beats static management policies depends on culture: in an agent-based model of the Care-Dare dilemma, cultural conformity can reverse the effectiveness of trust-based management. Agents update their behavior from the care and dare actions they receive, since people observe behavior but not latent dispositions. For consultants: trust-first strategies are not universal, and in highly conformist cultures the sequencing of care and challenge deserves explicit testing.
Is informality a better ‘fit’ in small and micro businesses? An example of training in the management of workplace conflict
International Small Business Journal Researching Entrepreneurship — University of Westminster · University of Sheffield
Online managerial training promoting proactive, collaborative responses to workplace problems produced positive shifts in conflict management styles and improved staff experience in micro and small firms, where employment relations are personal and informal. Strengthening everyday managerial interactions worked with the grain of informality rather than against it. For consultants: in small firms, upgrading informal conflict handling beats imposing structured procedures designed for large organizations.
Privacy Challenges of AI-Enabled Monitoring in the Workplace
Journal of the Australian Library and Information Association — Deakin University · Adelaide University
Employees at a large Dutch financial organisation showed general acceptance of AI-enabled monitoring when it serves learning, career development, and more effective workforce management. How monitoring is practiced and experienced determines whether it reads as support or surveillance. For consultants: purpose and transparency decide whether monitoring lands as development or distrust, so anchor any people-analytics rollout to visible employee benefit.
Knowledge sharing and telework in the new workplace: The role of organizational support in shaping employee outcomes
German Journal of Human Resource Management Zeitschrift für Personalforschung — Griffith University · Hanoi University of Industry
Perceived organizational support is associated with knowledge sharing on AI-enabled platforms among teleworkers in the UK, which in turn links to higher job satisfaction and security and lower cheating behavior and turnover intention, with the support-to-sharing link stronger under transformational leadership and varying by telework arrangement. Support, not surveillance, moves remote knowledge flows. For consultants: if remote knowledge sharing is weak, check perceived organizational support and leadership style before blaming the tools.
Trust, business group verticality, and firm productivity
Journal of International Economics — Max Planck Institute for Innovation and Competition · Centre for Economic Policy Research
Business groups whose members trust each other more are more vertically structured, and more trustworthy affiliates sit in layers closer to the ultimate owner and show higher productivity, in a large European sample built from ownership chains. Trust shapes structure, and structure correlates with performance. For consultants: trust is an organizational design variable with measurable productivity correlates, useful evidence when clients dismiss culture as soft.
AI-coaching and well-being in sales: a theory of tools perspective
Journal of Personal Selling and Sales Management — Bournemouth University · University of Arkansas at Fayetteville
Sales research has fixated on what AI does to productivity; this paper builds the missing well-being agenda. Using the theory of tools, it distinguishes AI coaching that is ready-to-hand (transparently integrated), unready-to-hand (failing mid-task) and present-at-hand (analytically observed by the user), and sets out a research agenda on the lived experience of coached salespeople, with well-being outcomes that can run positive or negative. For consultants: when a client rolls out AI coaching for a salesforce, the measurement plan should include the experience of the coached, not just quota attainment; this framework supplies the vocabulary.
Commuting matters: Association between work travel and perceived health and well-being in Canada
Transportation Research Part D Transport and Environment — McGill University · University of Iowa
Canadian Mobility Survey data (N=6,671) links how people get to work with physical health, mental health and life satisfaction: cycling and walking beat driving on all three, rapid transit helps physical health, and the commute emerges as a wellbeing variable in its own right, not a neutral cost of employment. For consultants: hybrid design debates price the office but rarely the journey; the evidence says the commute itself belongs in the workplace-policy conversation.
The future of work in creativity and innovation: perspectives on on-site versus remote work policies and practices
Industry and Innovation — Harvard University · London School of Economics and Political Science
A debate synthesis by Roche, Choudhury, Larson and Ter Wal weighs the strongest version of each case: sustained co-presence enables knowledge spillovers, tacit exchange, serendipity and cultural cohesion, while mandates cost autonomy, trigger psychological reactance, favour homophily and risk mistaking presence for collaboration just as innovation goes distributed. For consultants: treat return-to-office as a design question, not a policy toggle; identify which co-presence mechanisms the client actually needs and buy those, not attendance.
Working-time reduction: from a weekly working-time reduction to a reduction of weekly working days?
Transfer European Review of Labour and Research — Université Paris Sciences et Lettres · Institut de Recherche Interdisciplinaire en Sciences Sociales
Boulin traces how the postwar standard working week was destabilised by non-standard hours and work intensification, and reads the four-day-week trials running in several countries, with positive results for employees and businesses, as a shift in the frontier: from reducing weekly hours to reducing weekly working days. For consultants: clients asking about the four-day week are really choosing between compressing hours and cutting days, and the health, coordination and coverage consequences differ.
Trapped in time: the problem of crip time in the workplace for chronically ill and disabled people
Disability & Society — University of Cambridge · University of York
Co-produced qualitative work with Black disabled workers living with sickle cell (two focus groups, 47 interviews) finds that uncertain temporality shaped disclosure and requests for reasonable adjustments, that time at work ran on ableist norms, and that participants who succeeded did so precisely by organising their time flexibly. The authors warn that focusing solely on flexibility to work is 'double think' when wider policy and legislation stay unchanged. For consultants: flexible-work design is also an inclusion instrument; audit whose time norms a schedule assumes before presenting flexibility as the accommodation.
Flexibility stigma in the UK and Germany post-pandemic: intersectionality of stigma across gender and parental lines
Community Work & Family — King's College London · National University of Singapore
Homeworking carries career penalties in both the UK and Germany after the pandemic, particularly for those working from home more often, in experimental vignette data from two national panels. Penalties vary across gender and parental status against a backdrop of ideal worker norms and breadwinning expectations. For consultants: hybrid policies on paper do not remove stigma in practice, so examine how promotion decisions treat frequent homeworkers before declaring flexibility real.
Remote work frequency and loneliness among United Kingdom employees: the moderating role of active and passive social media use
Frontiers in Psychology — Peking University · Beijing Normal University
Higher remote work frequency is associated with greater loneliness among UK employees, and active social media use (posting content) buffers the association while passive scrolling does not, in fixed-effects models on 7,601 individuals from the UK Household Longitudinal Study. What remote workers do online matters more than time spent online. For consultants: counter remote loneliness with active, expressive digital participation, not just more channels to browse.
Using emoji reduces dropout of remote workers: a causal analysis on GitHub
Humanities and Social Sciences Communications — University of Arizona · University of Maryland, College Park
Including emojis in at least one post cut remote workers' dropout risk in the following year by half on GitHub, in a causal analysis using propensity score methods, and higher emoji intensity lowered dropout further. Emotional signals compensate for the delayed feedback and reduced emotional support of asynchronous text. For consultants: in async collaboration, small emotional cues are retention infrastructure rather than decoration, worth normalizing in written culture.
Caring Flexibly? Working Conditions and the Supply of Adult Care
Health Economics — London School of Economics and Political Science
The UK expansion of the Right to Request Flexible Work raised the likelihood that men provide informal care within the household by 1.3 percentage points, alongside greater control over working hours, while reducing high-intensity caregiving among women. Flexibility redistributed care work as well as schedules, in gendered ways. For consultants: flexible work policies reshape life outside work, and anticipating those second-order effects belongs in policy design advice.
Gender and Racial Disparities in the Earnings Effects of Remote Work Before and During the COVID-19 Pandemic: The Case of Brazil
Population Research and Policy Review — Universidade Estadual de Campinas (UNICAMP) · Columbia University
Working from home carried an earnings penalty in Brazil before the pandemic, and during it the penalty shrank for men, especially Black and Brown men, while remaining high for women; White women became likelier to work remotely but less likely to be promoted to managerial positions when doing so. Remote work redistributes penalties unevenly by gender and race. For consultants: remote arrangements can quietly widen pay and promotion gaps, so equity monitoring belongs inside any hybrid policy.
Factors Shaping Attitudes Towards Remote Work in the Czech Republic
Ekonomické rozhľady – Economic Review — New York University · Prague University of Economics and Business
Full-time employees in remote-friendly Czech companies show a significantly stronger inclination toward remote work than part-time colleagues, while age and gender show no statistically significant effects, in survey data from 227 respondents. Employment status, an understudied predictor, beats demographics. For consultants: do not assume generational preferences drive remote work demand, since contract type and work situation predict it better.
Sustaining Employee Engagement and Wellbeing in Hybrid Work: Strategic Perspectives for HRM Professionals
Merits — Centre for Global Health Research · University College London
Employee wellbeing and engagement in hybrid work environments are declining, and a synthesis of 78 studies grounded in the Job Demands-Resources model identifies the determinants HRM professionals can act on. Hybrid work needs deliberate design rather than default arrangements. For consultants: treat hybrid wellbeing as a design problem with identifiable determinants, because engagement erodes when arrangements are left to inertia.
How Artificial Intelligence Shapes Human Narratives
Journal of Language and Social Psychology — University of California, Los Angeles
A psychological account of narrative agency formalizes how groups learn from and influence narratives in digital contexts, and argues AI outputs only cause change in the world after passing through a human's narrative agency, making agency preservation the core design requirement for human-AI systems. For consultants: a peer-reviewed frame for the argument that AI deployments which bypass people's sense of authorship fail, and that alignment starts with who gets to tell the story.
AI and the Future of Accessible Work
Communications of the ACM — University of California, Irvine · Northwestern University
Experts in accessibility, organizations, AI and disability policy identify six key consequences of AI for accessible work, and insist they are structural, organizational, social and emotional, not merely technical: the same systems that offer real-time captioning, alt-text and writing support can create new barriers, leaving disabled workers to simultaneously depend on and deplore the tools. For consultants: accessibility belongs inside the AI deployment assessment itself, because the consequences land at the level of work design and organizational structure, not just the interface.
Shaping Collaborations with Algorithms: How Agency and Heterogeneity Criteria Influence Team Formation and Outcomes
Proceedings of the ACM on Human-Computer Interaction — University of Notre Dame · Northwestern University
Algorithms already shape who finds whom on networking platforms, co-founder matching and workplace collaboration tools. This study frames the design tension directly: optimise team composition for organizational goals, let people choose freely, or nudge in between, with each choice deciding who gains access to collaborators, opportunities and networks. For consultants: team-formation software is a governance decision wearing a UX costume; ask who set the objective function before a client adopts it.
Personality pairing improves human–AI collaboration
Proceedings of the National Academy of Sciences — Johns Hopkins University · Massachusetts Institute of Technology
A large preregistered experiment paired 1,258 people with AI agents prompted to exhibit different Big Five personalities; the human-AI teams produced 7,266 real display ads evaluated by 1,168 independent raters, plus a field experiment on X with nearly 5 million impressions. Human and AI personalities each shaped ad quality and teamwork, and the pairing itself mattered: extraverted humans matched with conscientious AI produced the lowest quality ads. For consultants: 'which model' is the wrong granularity; configuring an agent's disposition to the person it works with is a design variable with measurable output effects.
The Impact of Generative AI on Collaborative Open-Source Software Development: Evidence from GitHub Copilot
Information Systems Research — The University of Texas at Austin
Using GitHub's proprietary usage data, this study finds Copilot raised project-level contributions by expanding participation and individual output, then charged part of the gain back in coordination, review and integration time. Net integrated code still rose, but unevenly: core developers captured more of the benefit while peripheral contributors' work took relatively more effort to absorb. For consultants: budget the coordination bill when a client projects AI productivity gains, and expect the benefits to concentrate in people who were already central.
Is Our Future Colleague Even Human? Advancing Human–AI Teamwork from an Organizational Perspective
Journal of Organizational Behavior — Eindhoven University of Technology · Amsterdam University of Applied Sciences
AI reshapes collective work processes, not only individual judgments: design choices such as human-likeness and gendered cues shape perceptions, and AI alters team decision-making, trust, and stress. A sociotechnical framework distinguishes where AI sits in the social system of work and how much agency it exercises. For consultants: place each AI deployment on two axes, position in the team and degree of agency, before advising on integration and training.
How leaders build employee trust in artificial intelligence: voice opportunities, humility, and trust transfer
Journal of Business Research — Singapore Management University · King Fahd University of Petroleum and Minerals
Leader-provided voice opportunities reduce employees' negative affect about AI-related work experiences, raising perceived leader trustworthiness and, in turn, trust in AI, and the effect depends on leader humility signaling that concerns will be received openly. Trust in AI is social, not just technical. For consultants: when AI rollouts stall on skepticism, coach leaders to open voice channels and show humility instead of adding another accuracy briefing.
Fighting Fire with Fire: Infusing Artificial Intelligence into Peer Review to Sustain Quality Scholarship
Management Science — University of California, Davis · Stanford University
Large language models deployed as structured, trained, frontline reviewers can relieve strained journal review capacity, provided AI evaluations remain transparent and contestable within a human-in-the-loop oversight framework, compared through a mathematical model of journal operations. Efficiency, accuracy, and accountability are designed together, not traded off. For consultants: the template travels beyond publishing, so consider AI as an accountable first-pass reviewer wherever expert review is a client bottleneck.
Explainable AI: learning from the learners
Nature Communications — University of Michigan · University of Washington
Explainable AI combined with causal reasoning and domain validation can expose model-internal decision processes, generate candidate mechanistic hypotheses, and support trust and accountability in high-stakes applications. Opaque systems become partners for discovery, design, and safer use when their reasoning is opened up. For consultants: push clients past black-box adoption, because explainability is what turns AI from an oracle into a colleague whose judgment can be audited.
Exploring Large Language Model‐Based Intelligent Agents: Definitions, Methods, and Prospects
Wiley Interdisciplinary Reviews Data Mining and Knowledge Discovery — Chinese University of Hong Kong, Shenzhen · Peking University
LLM-based agents with advanced reasoning, planning, and tool-use capabilities are moving from theoretical constructs to practical applications, from autonomous task assistants to multi-agent simulations of social and economic systems, organized here into a framework deconstructing single-agent and multi-agent architectures. Natural language becomes the universal interface for cognition and coordination. For consultants: agent systems will soon sit inside client workflows, and the single-agent versus multi-agent distinction is the first design question to master.
Artificial intelligence in government: why people feel they lose control
Journal of European Public Policy — LMU Klinikum · Ludwig-Maximilians-Universität München
People confronted with algorithmic delegation in government judge it along three tensions: assessability (can decisions be understood), dependency (can the delegation be reversed), and contestability (can decisions be challenged), with early functional gains obscuring long-term risks to legitimacy. Control, not efficiency, is where acceptance breaks. For consultants: the same three questions apply inside companies, because employees who cannot understand, reverse, or challenge algorithmic decisions will feel they have lost control.
Performance gift-giving and incentive systems: Conditions and effects of peer helping in a sales team
Management Accounting Research — Copenhagen Business School
A Copenhagen Business School case study documents sales reps who hit their individual targets gifting surplus performance to teammates still short of theirs. What looks like collusion, inflating bonus payouts and distorting individual attribution, also stabilised the incentive system and strengthened team dynamics. For consultants: before tightening a client's anti-gaming controls, check whether the workaround is doing the cohesion work the formal system forgot.
Human capital profile similarity as a sorting mechanism in mergers and acquisitions
The Journal of Technology Transfer — IMT School for Advanced Studies Lucca · Bocconi University
Acquirers pair with targets whose human-capital profiles resemble their own, measured the year before the deal across 188,716 realised merger and acquisition pairs, consistent with absorptive-capacity theory: similar workforces are easier to absorb. Workforce alignment operates as a sorting mechanism before the deal, not just an integration issue after it. For consultants: in M&A advisory, profile both workforces early, since human-capital similarity shapes both match formation and post-acquisition knowledge absorption.
Employee Ownership as local economic resilience: Evidence from Scotland
Local Development & Society — University College Cork · Rutgers, The State University of New Jersey
Founders of locally embedded SMEs in Scotland choose employee ownership over trade sale, private equity acquisition, or closure to preserve organizational culture, retain skilled employment, and keep decision-making local, and the transition redistributes entrepreneurial agency to employees who assume governance and strategic responsibilities. Employee ownership supports local value retention and resilience. For consultants: succession is an opening for workplace democracy, and employee ownership sells to founders as continuity rather than ideology.
The employee share ownership participation decision: An integrative review and agenda for future research
Human Resource Management Review — The University of Melbourne · Institute on Governance
Employee share ownership participation is really three interrelated decisions (buy, contributions, holdings), and a review of 48 publications across disciplines maps the antecedents and mechanisms behind each within employee stock purchase plans. Prior research on attitudes and firm performance says little about why employees join or how long they stay. For consultants: when a client's share plan has low uptake, diagnose the three decisions separately instead of treating participation as one yes-or-no choice.
The Impact of FAS 123R on Employee Retention
Accounting Perspectives — University of Waterloo · University of Toronto
After FAS 123R reduced the use of employee stock options, employee turnover rose relatively more in metropolitan areas with lower noncompete enforceability and in states with greater R&D expenditure and no inevitable disclosure doctrine laws. Options were doing real retention work, visible once the accounting change cut their use. For consultants: broad-based equity is a retention instrument whose removal shows up in quits, evidence for clients weighing whether ownership stakes are worth the cost.
¿Inteligencia artificial cooperativa?
Hipertextos — Harvard University · Scuola Normale Superiore
Cooperatives, with nearly two centuries of history and a billion participants, are presented as laboratories capable of articulating democratic, transparent, common-good AI, though they face structural tensions including self-financing dependence, market co-optation risk, and internal reproduction of hierarchies. The central question of AI governance is framed as ownership and control of infrastructure, not ethics or regulation. For consultants: who owns the AI stack is becoming an organizational design question, and the cooperative option deserves a place in client conversations about technological sovereignty.
Reimagining ownership and governance for carbon dioxide removal (CDR): exploring existing alternative models to advance equity and justice
Carbon Management — University of British Columbia · Carbon180
Alternative ownership arrangements, including cooperatives and common-pool models drawn from agriculture, energy, mining, aquatic resources, and water infrastructure, offer routes to scale carbon dioxide removal more justly than the private-sector logics that currently dominate. Governance models transfer across sectors with structural similarities. For consultants: sectors being built from scratch give alternative ownership models their fairest test, and the repertoire of cooperative and commons designs is already documented.
Stress, resilience, and recovery: assessing a physical restorative space within a multimodal support program for hospital workers
Frontiers in Organizational Psychology — Johns Hopkins University · Johns Hopkins Medicine
Johns Hopkins Hospital measured its RISE Restorative Space with badge-swipe data and employee reports across four months: a cluster of rooms for relaxation and peer support in a high-stress clinical environment, assessed as a workplace design intervention rather than an individual resilience program. For consultants: recovery infrastructure is an organizational design choice with measurable uptake; it shifts the burnout conversation from fixing people to fixing conditions.
Workplace stress and healthy aging: assessment strategies, research hypotheses, and pilot insights from the Semmelweis study
GeroScience — Semmelweis University · Hungarian Institute of Occupational Health
Workplace stress assessment is being embedded into a large longitudinal occupational cohort, the Semmelweis Study, to test how stress shapes the biological, psychological, and social determinants of healthy and unhealthy aging in a university workforce, starting with a pilot in a high-risk subgroup. Sustainable pace becomes measurable as an aging outcome, not just a satisfaction score. For consultants: linking chronic workplace stress to how people age gives wellbeing work a longer horizon than engagement surveys capture.
Burnout as Social Information: How Coworker Burnout Level and Coworker Burnout Consistency Interact With Mastery Goal Orientation to Affect Work Performance
Stress and Health — Wilfrid Laurier University · Hong Kong Baptist University
Coworker burnout acts as social information: when most coworkers in a group show high burnout, employees read the cue, reduce work effort, and perform less well. The contagion is informational as much as emotional, filtered through what effort seems worthwhile in that environment. For consultants: burnout spreads through what teams signal about effort norms, so address the shared signal rather than only the individuals showing symptoms.
Positive Psychology in the Workplace: Psychological Capital, Flourishing, Leadership, and Employee Well-Being in Contemporary Organizations
Administrative Sciences — National Centre for Social research
Psychological capital, emotional intelligence, mindfulness, psychological safety, self-determination, and positive leadership contribute to employee wellbeing, flourishing, and organizational effectiveness, synthesized into an integrative framework from recent research in positive organizational psychology. The constructs form a coherent system rather than a menu of perks. For consultants: wellbeing programmes gain traction when built on these evidence-based psychological resources instead of isolated benefits.
‘Holistic exploration’ in coaching: practitioner accounts of how an integrative workplace coaching programme was enacted and understood
Coaching An International Journal of Theory Research and Practice — University College London · State Administration of Cultural Heritage
Coaches in a UK therapeutic coaching programme for business owners described their shared practice as holistic exploration, an orientation spanning performance, development, wellbeing, and mental health needs rather than a fixed model or set of techniques. Integrative workplace coaching is defined by stance more than toolkit. For consultants: coaching that engages wellbeing alongside performance is already how practitioners work, so contracting should name that breadth explicitly.
Unveiling the Role of Supervisor Workaholism: Impact on Idiosyncratic Deals and Employee Promotability
Human Resource Management Journal — Miami University · Peking University
Supervisor workaholism is negatively related to granting schedule flexibility and location flexibility i-deals, while task i-deals link positively to employee promotability, in data from 414 employees and 54 supervisors in China. Whether flexibility gets granted depends on the boss's own relationship to work. For consultants: idiosyncratic deals flow through supervisors' work identities, so flexibility policies fail quietly under workaholic managers unless the pattern is surfaced.
The protective role of perceived support, job recognition and availability of workplace resources in the relationship between burnout and suicidality among health and social care professionals in Switzerland
Frontiers in Public Health — HES-SO Fribourg · University Psychiatric Hospital
Burnout among health and social care professionals in Switzerland is significantly related to suicidality, and perceived private support, workplace support, availability of workplace resources, and job recognition mediate that relationship, in data from 2,388 professionals. Support and recognition sit on the pathway between exhaustion and its worst outcomes. For consultants: recognition and resource availability are measurable protective factors worth auditing in any client workforce under strain.
Balancing Teacher and Student Agency: Co-Orchestration Tool Design Supporting Real-Time Dynamic Pairing
Proceedings of the ACM on Human-Computer Interaction — Carnegie Mellon University · University of Washington
Design research at Carnegie Mellon on classroom co-orchestration tools asks who controls the real-time pairing of learners, teacher or student, across the before, during and after of each pairing. The premise travels beyond schools: respecting user agency is what sustains trust in human-AI orchestration. For consultants: any AI that assigns people to work needs an explicit answer to where the human override sits.
Integrating pedagogy for autonomy with exploratory practice in English for academic purposes
Journal of English for Academic Purposes — Queen Mary University of London · University of Warwick
Integrating learner autonomy with exploratory practice in a UK pre-sessional English course left students satisfied with working together, exercising their agency, and investigating their own epistemic needs, evidenced across reflective diaries, presentations, and feedback a year later. Agency was built into the course rather than added on. For consultants: adult learning designs that hand investigation over to learners produce engagement that outlasts the course, a pattern that transfers to corporate development programmes.
From problems to challenges: a genealogical literature review of challenge-based learning and future directions
European Journal of Engineering Education — Department for Education · University College London
Challenge-based learning discourse in engineering education combines innovation, urgency, external relevance, and student agency while obscuring the histories, conflicts, and power relations through which challenges are constituted; external partnerships rose from 9% of studies in the foundation phase to more than 77% in the expansion phase. Authenticity increasingly means stakeholder participation. For consultants: challenge framing energizes learners but smuggles in whose problems count, a question worth asking when designing challenge-based development for clients.
From andragogy to heutagogy: a scaffolded pathway for self-directed neurosurgical learning
Journal of Clinical Neuroscience — University of Edinburgh · Sher-i-Kashmir Institute of Medical Sciences
A three-stage pathway from foundations through andragogy to heutagogy maps increasing learner autonomy onto training milestones and continuing professional development, built for neurosurgical training where teacher-centred models no longer produce the adaptive experts needed. Self-determined learning is staged, not assumed. For consultants: scaffold autonomy progressively, because even in the most expertise-dense professions learners graduate into self-direction stage by stage.
Scaffolding in the engineering lab: Organic and self-directed varieties of scaffolding
Mind Culture and Activity — University of North Texas · The University of Texas at Arlington
An engineering lab advisor scaffolded students' learning without knowing or intending the strategy, an organic scaffolding made possible by continuous encouragement toward independent research, a nonhierarchical collaborative environment, and students inclined to self-directed learning; problem-solving interactions were co-constructed and student-initiated. Structure emerged from culture, not curriculum. For consultants: nonhierarchical environments generate their own developmental support, an argument that flat structures and learning are complements.
Investigating the reciprocity between daily emotions and perceived learning progress in self-directed learning
Cognition & Emotion — Nanyang Technological University · Hong Kong Polytechnic University
Daily positive emotions and perceived learning progress reinforced each other reciprocally across a roughly 10-week self-directed learning project involving 321 adolescents, while rises in daily negative emotions predicted worse subsequent outcomes. Emotion is a working part of self-directed learning, day by day. For consultants: in self-paced development programmes, track how learners feel daily, since mood and progress feed each other in both directions.
Generative AI, Cognitive Offloading, and Learner Agency in Higher Education: A Scoping Review
Behavioral Sciences — Tsinghua University · Harbin Institute of Technology
Generative AI in higher education supports writing, feedback, problem solving, and research tasks while raising concerns about cognitive offloading and learner dependence, mapped in a PRISMA-guided scoping review of peer-reviewed literature from 2022 onward. The tension between offloading and learner agency is now a defined field of evidence. For consultants: AI-supported learning trades effort for speed, and preserving learner agency requires deliberate design, in classrooms and corporate academies alike.
Self-directed learning readiness in preclinical medical education: Patterns, variation, and academic outcomes from a mixed longitudinal and cross-sectional study
BMC Medical Education — Sam Houston State University · Western University
Self-directed learning readiness, spanning learning motivation, planning and implementation, self-monitoring, and interpersonal communication, varies across the preclinical medical curriculum and relates to academic performance, in data from 434 unique students across three cohorts. Readiness for autonomy is multidimensional and measurable. For consultants: readiness for self-direction can be measured across distinct components, a diagnostic worth running before redesigning development around learner autonomy.
Wage growth of less educated workers: the importance of social skills
Fiscal Studies — INSEAD · Collège de France
Matched UK employee-employer data shows that social skills (teamwork and communication with co-workers) drive wage growth for workers with less formal education: in social-skill tasks a significant fraction of less educated workers enjoy wage progression with tenure, with higher returns in firms with more educated co-workers. The authors' model treats social skills as initially opaque to both employee and employer, learned about over time through complementarity with colleagues. For consultants: for workers without credentials, team composition is development policy; mixing them with educated co-workers is where the skill, and the wage, gets revealed.
Age-group perspectives on large language models in the architecture, engineering, and construction industry: usage patterns, adoption expectations, trust, and future outlook
Construction Management and Economics — New York University · New Jersey Institute of Technology
An age-stratified survey of 91 architecture, engineering and construction professionals finds younger and middle-aged practitioners report far higher LLM familiarity and use than older colleagues, yet perceived usefulness among actual users is similar across age groups, and expectations diverge on 11 of 12 assistant capabilities. The adoption gap is about exposure, not payoff. For consultants: the generational gap closes on use, so seed usage across cohorts instead of writing older workers off.
Determinants of AI acceptance in the workplace: Evidence from eleven European countries
Computers in Human Behavior Reports — Eötvös Loránd University · Harvard University
Probability-based panel data from 5,250 employed adults in eleven European countries (ESS CRONOS3, fielded May to June 2025) shows that acceptance of AI at work is strongly role-dependent: the study disaggregates acceptance by the hierarchical role assigned to AI, from assistant to supervisor, and finds a single 'attitude to AI' number hides the real structure. For consultants: survey the client workforce by AI role, not by tool; acceptance of an assistant says little about acceptance of an algorithmic boss.
Associations of artificial intelligence use at work with psychological distress, loneliness, and life satisfaction among employed US adults
Preventive Medicine Reports — New York University
Among employed US adults, 51.7% use artificial intelligence at work, most commonly for searching for information or technical help (35.8%) and writing communications (31.6%), and the number of AI-assisted tasks shows measurable associations with psychological distress, loneliness, and life satisfaction in nationally representative data. AI use at work is now a majority experience with psychosocial correlates. For consultants: AI adoption is a wellbeing variable as well as a productivity one, so track psychosocial outcomes alongside usage metrics in client rollouts.
Human Resource Development’s Role in AI-Enabled Workforce Transformation
New Horizons in Adult Education and Human Resource Development — Oakland University · The University of Texas at Tyler
Human resource development professionals are positioned as the bridge between digitalization and future workforce competencies, guiding and moderating AI integration amid talent shortages, skill gaps, ethical concerns, and continuous upskilling needs, with proposed strategies covering an AI-ready culture, systematic AI training, and ethical policies. Workforce transformation is framed as an HRD project, not only a technology project. For consultants: culture, training, and ethics policy form the practical AI agenda to bring clients whose plans stop at tool procurement.
From Blueprint to Black Box: How Generative Artificial Intelligence Transforms the Artistic Workflow
British Journal of Industrial Relations — Cornell University · New York State School of Industrial and Labor Relations
AI's effects on concept artists depend on worker involvement in technology deployment: involvement produces human-AI complementarity and stronger task-technology fit, while exclusion leads to skill substitution, work intensification, fragmented workflows, and heightened anxiety around technology. Occupational identity erodes when workers are shut out of implementation decisions. For consultants: worker participation in AI deployment is the variable separating augmentation from degradation, a direct design principle for client rollouts.
Techies and Firm‐Level Productivity
International Economic Review — National Bureau of Economic Research · University of Virginia
Techies (engineers and other technically trained workers) raise firm-level productivity beyond the contribution of R&D workers, extending to ICT and other technical staff, in French administrative data; in nonmanufacturing firms the effect operates mostly through ICT and other techies. Technical talent is tied to innovation, technology adoption, management, and diffusion inside firms. For consultants: productivity gains ride on embedded technical staff rather than formal R&D alone, so map where techies sit in a client's structure.
Can the fourth industrial revolution solve the productivity problem?
The Economic and Labour Relations Review — St George's, University of London · University of Manchester
Whether AI-driven productivity growth benefits workers depends on who captures the gains: zero-sum productivity accrues to capital owners while positive-sum productivity is broadly shared, and corporations' tendency not to share gains is among three factors undermining positive-sum outcomes. Trillions in projected GDP say nothing about distribution. For consultants: productivity pitches for AI adoption should come paired with explicit gain-sharing design, or the workforce has little reason to cooperate.
Sakura and Shifting Tides: Exploring the Deskilling and Upskilling Experiences of Indonesian Nurses in Post-COVID-19 Japan
Caring Indonesian Journal of Nursing Science — University of Indonesia · Kyoto University
Indonesian nurses working in post-COVID Japan describe deskilling alongside deliberate upskilling and reskilling strategies, workplace changes, cultural and professional adaptation, and career transition planning, across interviews with 41 nurses in hospitals and long-term care facilities. Skill trajectories under labor migration run in both directions at once. For consultants: mobile workers actively manage their own reskilling, and retention advice should account for skill loss and gain happening simultaneously.
An integrated AI literacy framework for higher‑education teachers: a critical review and conceptual synthesis
Cogent Education — University College London
Teacher AI literacy shows significant gaps in ethical understanding, innovative use of AI, and institutional support, according to a critical interpretive synthesis of 23 peer-reviewed studies that builds a holistic competency model extending beyond foundational knowledge to evaluating, creating, and deploying AI tools. Literacy is a layered competency, not a single skill. For consultants: AI literacy frameworks built for educators translate to any expert workforce, covering knowledge, application, evaluation, creation, and the institutional support layer clients usually forget.
Wage Adjustment in Efficient Long-Term Employment Relationships
American Economic Journal Macroeconomics — University of Edinburgh · Federal Reserve Bank of Cleveland
A model of efficient long-term employment shows wages adjust only sporadically because renegotiation is costly, hiring wages have surprisingly little influence on later pay and allocation (a near-memorylessness property), and recruitment and retention bonuses do crucial work in sustaining the relationship. For consultants: the starting salary matters less than clients think; the design of renegotiation moments and retention mechanisms is where long-term deals are actually won.
Communicating Corporate Culture in Labor Markets: Evidence from Job Postings
Management Science — Hong Kong University of Science and Technology · Harvard University
Machine-learning measurement of five core values in job postings shows that culture information attracts job seekers, with higher worker inflows, and the effect strengthens when seekers value culture and face frictions learning about it elsewhere. Culture in the posting is also associated with more interviews converting to offers and more offers accepted. For consultants: the job posting is a sorting device, not decoration; writing the culture into it moves the funnel at both ends by matching culturally aligned candidates to the firm.
Experimental Analysis of Gender Bias in Hiring within STEM Fields
Universitätsbibliothek der LMU — Technical University of Munich
Students evaluating systematically varied CVs in STEM showed no consistent bias in favor of male candidates, but strong male applicants received higher salary suggestions while female candidates were more likely to be offered professional support, with effects most pronounced among Eastern European and older participants. Bias emerges in interactions between candidate gender and resume strength, not as a blanket preference. For consultants: hiring bias hides in salary and support decisions even when competence ratings look fair, so audit each decision separately.
Scientific Talent Leaks Out of Funding Gaps
The Review of Economics and Statistics — University of Toronto · United States Census Bureau
Interruptions of at least 30 days in NIH grant renewals increase nonemployment among personnel in single-grant labs by 3 percentage points, or 40%, and earnings fall 20% among those who remain employed in the US, with trainees especially affected. Short funding gaps bend career trajectories for good. For consultants: continuity of funding is a talent retention issue, because even brief institutional gaps push skilled people out of the field entirely.
From motorized projectors to generative AI: cultural conflicts over labor automation in Hollywood
AI & Society — University of Southern California · University of Pennsylvania
Whether a technology counts as automating is culturally constructed rather than technically determined: three Hollywood cases (motorized projectors in the early 1900s, computer-based editing in the mid-1980s, screenwriting software in the late 1980s) show competing narratives among workers and other actors shaping the outcome, which reframes generative AI as less unprecedented than claimed. Automation fights are narrative fights. For consultants: who defines a tool as replacing versus augmenting skill shapes how the labor conflict resolves, and that framing can be worked on deliberately.
Effects of Uber and Lyft on jobs, wages and GDP
Nature Cities — University of Oxford · Carnegie Mellon University
Uber and Lyft entry across 167 US service regions increased regional GDP per capita and seasonal, temporary, or otherwise intermittent employment per working-age population, with no statistically significant effects on overall employment or earnings. Gig platforms added intermittent work at the margin rather than transforming the labor market. For consultants: platform work grows the contingent edge of employment, relevant context when clients weigh gig-style staffing against standard roles.
Nonparametric estimation of matching efficiency and elasticity in a spot gig work platform: 2019–2023
Journal of the Japanese and International Economies — Waseda University · The University of Tokyo
Comparing Japan's spot gig platform Timee with the public employment service, matching efficiency ran substantially higher on the private platform, especially after 2022, for short-term jobs with minimal educational or experience requirements. Credential-light labour is clearing through faster private channels. For consultants: the flexible end of the labour market is being re-intermediated by platforms, and clients hiring through slower public channels are competing against that clock.
Beggars cannot be choosers: Labor market tightness, hiring standards, wages, and hiring costs
Labour Economics — Maastricht University · Federal Institute for Vocational Education and Training
In tighter labor markets, firms do not raise entry wages; they lower hiring standards, taking workers with lower general ability relative to their existing workforce, and absorb higher post-match costs, which already account for 84 percent of total hiring expenditures. Adjustment happens through standards and onboarding investment, not pay. For consultants: when clients cannot hire, the realistic play is relaxing credential requirements and budgeting for a longer adaptation period, exactly what firms in tight markets already do.
Does Weaker Employment Protection Lower the Cost of Job Loss?
The Journal of Human Resources — University of Essex · IZA - Institute of Labor Economics
Weakening employment protection in Italy's 2012 reform raised the cost of job loss: displaced workers faced lower re-hiring earnings and lower re-employment probabilities, with penalties deepest for labor market outsiders such as young workers and people in poorer regions. Less protection made displacement more expensive rather than easing re-entry. For consultants: claims that flexibility helps workers back into jobs did not hold in this evidence, useful context when advising on workforce restructuring.
Unequal Progress and Barriers for Women in Economics
Reviews of Economic Literature — London School of Economics and Political Science · SOAS University of London
Women now constitute 26% of published economists but only 5% of the top percentile of research influence, with recent progress paradoxically concentrated at that highest level as institutional norms shift. Barriers operate through career choices, gender biases in evaluation, and workplace environments. For consultants: representation gains at entry do not propagate upward on their own, and evaluation and environment redesign is where the leak actually sits.
Reimagining Hospital Medicine for Long-Term Sustainability
Journal of General Internal Medicine — University of Wisconsin–Madison · Wisconsin Division of Public Health
High burnout and turnover make hospital medicine hard to sustain as a career, and focus groups with hospitalists identify factors that promote or impede career longevity along with job-related and age-related challenges facing older clinicians. Career sustainability is a design problem the field is only beginning to study. For consultants: longevity factors differ by career stage, so employment deals should stay renegotiable as workers age rather than being fixed at hiring.
Age and generational differences in anthropomorphism and trust in large language models
Frontiers in Cognition — Google (United States) · University of California, Davis
Gen Z showed consistently lower anthropomorphism, lower trust ratings, and lower accuracy ratings for LLM responses than older generations across 1,485 participants, while anthropomorphic voice cues increased perceived human-likeness and accuracy uniformly across age groups. The youngest users are the most skeptical of these systems. For consultants: do not assume digital natives trust AI most, because younger employees may need different adoption arguments than their managers expect.
Age, gender, and score distributions of moral foundations
PLoS ONE — University of Edinburgh
Binding moral foundations (proportionality, loyalty, authority, purity) rise consistently with age in US and UK samples, accumulating to differences of roughly 0.75 standard deviations between youngest and oldest adults, while preference for equality declines with age and harm/care stays stable. Value gaps between generations are real, gradual, and patterned. For consultants: intergenerational friction over authority and fairness at work tracks measurable moral differences, so design collaboration norms that translate across them rather than dismissing the gap as stereotype.
Surgical residents’ career interests in transplantation surgery in Germany – a nationwide survey
Transplant International — Heidelberg University · University Hospital Heidelberg
A quarter of surveyed surgical residents in Germany would consider transplantation surgery as a subspecialty, motivated by personal interest, patient care, and career development, yet transplant-specific education is occasional or absent even though more than 80% want it, and the main barriers are structural. The next generation's interest exists; the pathway does not. For consultants: when a field struggles to attract young talent, check whether the deal fails on interest or on structure, because the fixes differ completely.
“Sanctified Slack?”: How Religiosity Shapes the Link Between ‘Quiet Quitting’ and Psychological Distress
Journal of Religion and Health — Baylor University · University of Toronto
Quiet quitting is positively associated with psychological distress among US workers: not putting in extra effort and unwillingness to work harder both correlate with worse mental health, though beliefs in divine control buffer the link. Withdrawing effort is not psychologically free. For consultants: quiet quitting signals distress rather than contentment with less, so treat it as a mental health and engagement indicator instead of laziness.
Universities & business schools on the radar
Harvard University · MIT · Stanford University · University of Oxford · University of Cambridge · Caltech · Princeton University · UC Berkeley · Yale University · Imperial College London · ETH Zurich · UCL · University of Chicago · University of Pennsylvania · Columbia University · Johns Hopkins University · Cornell University · University of Toronto · University of Michigan · UCLA · National University of Singapore · Tsinghua University · Peking University · University of Tokyo · Kyoto University · EPFL · University of Edinburgh · Northwestern University · New York University · Duke University · Université PSL · Sorbonne Université · TU Munich · LMU Munich · London School of Economics and Political Science · London Business School · INSEAD · Bocconi University · HEC Paris · Copenhagen Business School · Erasmus University Rotterdam · University of Warwick · Carnegie Mellon University · The University of Texas at Austin · The University of Melbourne · University of British Columbia · Western University · University of Manchester