# AI and the Workforce Summary

##  Recap of the 2026 AI and the Workforce Conference at NYU Washington DC

![cohosts smiling at the close of conference](/sites/default/files/styles/medium_375px_wide/public/2026-08/26-0668-821.jpg.webp?itok=7n_pD49L)

On Friday July 10, 2026, NYU’s [Center for the Future of Management](https://www.stern.nyu.edu/experience-stern/about/departments-centers-initiatives/centers-of-research/center-future-management?_gl=1%2A1g2skza%2A_gcl_au%2AMTgxNjQ5NTA2OS4xNzc5MTEwNzYx), [Fubon Center for Technology, Business and Innovation](https://www.stern.nyu.edu/experience-stern/about/departments-centers-initiatives/fubon-center), and [Development Research Initiative](https://nyudri.org/) hosted a conference on AI and the Workforce at NYU’s Washington DC campus.

The conference brought together academic, industry, government, policy and labor leaders for a series of discussions about understanding and shaping AI's current and future impact on the workforce. The discussions revolved around what the research tells us, what businesses are seeing in the field, how education must evolve to meet the challenge, creative interventions and their impact so far, and the key policy and legislative imperatives. What follows is a brief recap of the topics covered by the panels throughout the day.

##### **Quantitative Measurement of Workforce Impacts**

The first panel of the day, moderated by [John Soroushian ](https://www.linkedin.com/in/john-babak-soroushian-6495b324/)(Americans for Responsible Innovation) featured presentations by five research scholars.

![chandar speaking at podium](/sites/default/files/styles/medium_375px_wide/public/2026-08/26-0668-225.jpg.webp?itok=26Gsf8ZP)

[Bharat Chandar](https://www.linkedin.com/in/bharatchandar/) (Stanford) reviewed emerging evidence on AI's labor market impacts, concluding that economy-wide job displacement remains limited but that younger workers in highly AI-exposed occupations appear to be experiencing slower employment growth. He emphasized that while multiple studies point toward reduced entry-level hiring in exposed occupations, additional work is needed to disentangle AI's effects from other macroeconomic forces such as interest rates, remote work, and educational composition.

![gimbel speaking at podium](/sites/default/files/styles/medium_375px_wide/public/2026-08/26-0668-234.jpg.webp?itok=nN63SGXR)

[Martha Gimbel](https://www.linkedin.com/in/martha-gimbel-03b2598/) (Yale Budget Lab) presented evidence suggesting that, despite widespread attention to generative AI, there is not yet clear evidence of large-scale labor market disruption. While AI-exposed occupations have experienced somewhat greater occupational churn, employment, wages, and unemployment have not diverged significantly from comparable occupations since the introduction of modern large language models, underscoring the importance of distinguishing AI effects from broader economic and industry trends.

![simon speaking at podium](/sites/default/files/styles/medium_375px_wide/public/2026-08/26-0668-256.jpg.webp?itok=xqLSIu4C)

[Lisa Simon](https://www.linkedin.com/in/lisaksimon/) (Revelio) demonstrated how new employer and worker datasets can be used to track AI's impact in near real time, finding evidence that AI-exposed occupations have experienced weaker employment growth and reduced entry-level hiring while firms actively adopting AI often continue to grow. She also argued that AI is reshaping the composition of work itself, with jobs increasingly emphasizing human-centered activities even as AI automates portions of existing tasks.

![manning speaking at podium](/sites/default/files/styles/medium_375px_wide/public/2026-08/26-0668-270.jpg.webp?itok=vUmySQ0l)

[Sam Manning](https://www.linkedin.com/in/sam-manning-9a600244/) (GovAI) introduced a framework for measuring workers' "adaptive capacity" to AI-induced displacement, focusing on which workers are best positioned to transition successfully if disruptions occur. His analysis suggests that although many highly AI-exposed workers also possess strong adaptive capacity, millions of workers—particularly in clerical and administrative occupations—face the dual challenge of high AI exposure and relatively limited ability to adapt.

![schubert speaking at podium](/sites/default/files/styles/medium_375px_wide/public/2026-08/26-0668-290.jpg.webp?itok=xbHqs_FJ)

[Gregor Schubert](https://sites.google.com/view/gregorschubert) (UCLA) argued that AI's most important labor market effects may emerge through career trajectories rather than immediate job losses, highlighting changes in promotion opportunities, career mobility, and lifetime earnings potential. Drawing on evidence from prior technology shocks and new research on AI firms, he suggested that AI may already be altering promotion dynamics even while aggregate employment effects remain relatively small.

##### **Industry Signals About Workforce Impacts**

![nagle at podium](/sites/default/files/styles/medium_375px_wide/public/2026-08/26-0668-536.jpg.webp?itok=mkz1od1t)

The second panel, moderated by [Frank Nagle](https://www.linkedin.com/in/frank-nagle/) (MIT, Microsoft, and the Linux Foundation), brought together [Joe Atkinson](https://www.linkedin.com/in/atkinsonjoe/) (PwC), [Francis Hintermann](https://www.linkedin.com/in/francis-hintermann-243a9a1/) (Accenture Research),[ Welby Leaman](https://www.linkedin.com/in/j-welby-leaman-7251012/) (Walmart), and [Laura McGorman](https://www.linkedin.com/in/laura-mcgorman-7996b346/) (Meta) to offer practitioner perspectives on how AI is reshaping entry-level work and broader workforce strategies across consulting, retail, and technology.

![atkinson and hintemann in discussion](/sites/default/files/styles/small_230px_wide/public/2026-08/26-0668-538_0.jpg.webp?itok=1wSO_7zW)

Building on the empirical research presented during the preceding session, the panel examined where companies are seeing growth in demand for workers, where hiring or staffing may be contracting, and which emerging changes are not yet visible in conventional labor market data.

![leaman and mcgorman on the panel ](/sites/default/files/styles/medium_375px_wide/public/2026-08/26-0668-532.jpg.webp?itok=0raLrRPM)

The discussion explored how organizations are balancing automation with the augmentation of workers, how these patterns differ across industries and between customer-facing and operational roles, and which skills employers increasingly seek from new hires. Looking ahead to the conference’s sessions on higher education and workforce development, panelists also considered how universities, flexible learning models, apprenticeships, employer-led training, and reskilling initiatives can better prepare both new entrants and incumbent workers for increasingly AI-enabled workplaces

##### **Preparing Talent for the AI Economy**

![moderator and panelists posing for the camera](/sites/default/files/styles/medium_375px_wide/public/2026-08/26-0668-650.jpg.webp?itok=dKt2AHiN)

Moderated by [Mark Kennedy](https://nyudri.org/team/mark-kennedy/?author_team=920), the keynote panel featured university presidents [Andrew Armacost](https://campus.und.edu/directory/andrew.armacost) (University of North Dakota), [Pam Eddinger](https://www.bhcc.edu/officeofthepresident/) (Bunker Hill Community College), and [Todd Saliman](https://president.cu.edu/bio) (University of Colorado) discussing how higher education must evolve to prepare talent for an AI-driven economy.

![armacost speaking on the panel](/sites/default/files/styles/medium_375px_wide/public/2026-08/26-0668-574.jpg.webp?itok=ZKYuNuNf)

Panelists emphasized that AI represents both an opportunity and a challenge for universities, requiring institutions to rethink teaching, assessment, governance, and lifelong learning. The discussion highlighted the importance of AI literacy across disciplines, experiential learning through internships and apprenticeships, and stronger partnerships between higher education, industry, and government.

![eddinger and saliman sitting on the panel](/sites/default/files/styles/medium_375px_wide/public/2026-08/26-0668-589.jpg.webp?itok=Q2dpMpQ2)

While AI can expand access and personalize learning, successful adoption will depend as much on institutional leadership and change management as on technology itself. The panel concluded that preparing talent for the AI economy is not simply an educational challenge but a national competitiveness imperative.

A summary of the panel is [**available here**](https://nyudri.org/publications/preparing-talent-for-the-ai-economy/).

##### **Workforce Development and Apprenticeships**

This panel featured a series of short presentations on workforce development and apprenticeships, followed by a panel discussion of the role of apprenticeships to aid workforce learning and development in the age of AI.

![impink and jabbari at podium](/sites/default/files/styles/medium_375px_wide/public/2026-08/screenshot_2026-08-06_at_9.57.51_am.png.webp?itok=L2Ma_uLk)

[Michael Impink](https://www.hec.edu/en/faculty-research/faculty-directory/faculty-member/impink-stephenmichael) (HEC Paris) examined how previous waves of technological change—from industrial machinery to information technology—reshaped work only after complementary organizational and workforce changes occurred. [Jason Jabbari](https://brownschool.washu.edu/faculty-and-research/jason-jabbari/) (Washington University St. Louis) presented research showing that apprenticeships and other work-based learning models can substantially improve labor market outcomes by strengthening employer connections and increasing earnings.

![colborn and mitchell at podium](/sites/default/files/styles/medium_375px_wide/public/2026-08/screenshot_2026-08-06_at_10.03.54_am_0.png.webp?itok=M2QHFNNI)

[John Colborn](https://www.linkedin.com/in/johncolborn/) (Apprenticeships for America) provided an overview of apprenticeship programs in America, including how they’ve evolved over time and some of the challenges they currently face. [Cierra Mitchell](https://www.apprenticeship.gov/about-us/national-office) (U.S. Dept of Labor) provided an overview of the U.S. Department of Labor's Registered Apprenticeship system, highlighting its five core elements—paid employment, structured on-the-job learning, classroom instruction, industry leadership, and nationally recognized credentials.

![moderator and panelists posing for the camera](/sites/default/files/styles/medium_375px_wide/public/2026-08/26-0668-728.jpg.webp?itok=-_SengEU)

The discussion focused on how AI is changing the nature of entry-level work and increasing the value of practical, experience-based learning. Panelists discussed whether an emerging "experience premium" makes traditional classroom instruction less sufficient on its own, and explored how apprenticeships, experiential learning, and stronger employer-education partnerships can provide workers with opportunities to build job-relevant skills while adapting to AI-enabled workplaces. The panel also examined how workers can continue to create value as AI becomes more capable, emphasizing augmentation over automation and the uniquely human capabilities that remain difficult to replicate.

**Industrial and Workforce Policy for the AI Era**

![moderator and panelists sitting on the panel](/sites/default/files/styles/medium_375px_wide/public/2026-08/26-0668-793.jpg.webp?itok=F4YhdIBK)

The final panel of the day featured a spirited discussion and debate about what industrial and workforce policy was needed for the AI era. It featured labor leader [Amanda Ballantyne](https://www.newamerica.org/people/amanda-ballantyne/) from New America, [Maureen Conway](https://www.aspeninstitute.org/people/maureen-conway/?gad_source=1&gad_campaignid=22192163519&gbraid=0AAAAA-vx0GHVfae9okVVmYsNLUjvBrn-x&gclid=Cj0KCQjwqPLOBhCiARIsAKRMPZp-xrV3_GqS6SHqacQAXKTMTd2yrpzCiytcL2X84DcTMecqS0K_-sgaAriVEALw_wcB), the long-time Executive Director of the Aspen Institute's Economic Opportunities Program, policy leader [Morgan Dwyer](https://www.linkedin.com/in/morgan-m-dwyer/) from OpenAI, and economist [Michael Strain](https://www.michaelrstrain.com/) who heads economic policy studies at the American Enterprise Institute.

![ballantyne and conway sitting on the panel](/sites/default/files/styles/medium_375px_wide/public/2026-08/26-0668-753.jpg.webp?itok=aWGAIy5W)

Building on the quantitative and qualitative evidence about how AI is reshaping work and the panel discussions about educational change, the panelists discussed what AI is actually doing, what the key current policy challenges raised by AI are and whether they were new, and what specific concrete solutions might exist for making sure the gains from the AI revolution are broadly shared.

![dwyer and strain sitting on the panel](/sites/default/files/styles/medium_375px_wide/public/2026-08/26-0668-782.jpg.webp?itok=IwDZSt6v)

Ballantyne framed the moment as a bit of a 'Wild West,' like being at the top of a rollercoaster without knowing if the ride will be good or the rails will come off. Strain explained how history suggests that the changes caused by AI may be slower than we anticipate, and made an impassioned case for why the country needs to focus more on the massive productivity gains that AI promises and the need to shift the narrative more towards abundance rather than displacement.

![ballantyne, conway and dwyer sitting on the panel](/sites/default/files/styles/medium_375px_wide/public/2026-08/26-0668-765.jpg.webp?itok=65_CgKA2)

Building on her extensive prior work about worker ownership, Conway reiterated the need for new systems that achieve ex-ante pre-distribution that leads to greater sharing and more decentralized capital ownership, rather than ex-post redistribution through systems like a universal basic income.

![conway and dwyer sitting on the panel](/sites/default/files/styles/medium_375px_wide/public/2026-08/26-0668-811.jpg.webp?itok=3KyiHskZ)

Dwyer, whose team played a key role in developing OpenAI’s recent Industrial Policy for the Intelligence Age blueprint, discussed how pre-distribution proposals like an AI sovereign wealth fund could work well in conjunction with other interventions that strengthened the safety net and ensured the availability and portability of benefits to non-traditional work arrangements.

While the panel debated and sometimes disagreed about the depth of the impending workforce disruption and whether current levels of US income redistribution was inadequate or excessive, they agreed about the need for urgent and bold industrial policy.

**Survey Results**

Following the conference, attendees were surveyed for feedback on which topics they felt were well represented, which topics they would like to see more of, and their overall impressions of the conference.

Respondents expressed the strongest interest in expanding programming focused on industrial policy and labor interventions alongside empirical evidence about AI's labor market impacts, while relatively few felt any topic was overrepresented. Frequently requested additions included AI's effects on entrepreneurship, business formation, macroeconomic outcomes, the resilience of unemployment insurance and other social safety nets under larger-scale labor displacement, AI safety and security, AI applications in the public sector, and the broader societal implications of increasingly capable AI systems.

Multiple comments suggested inviting more practitioners from AI companies—particularly product leaders and developers—to discuss expected advances in model capabilities and practical applications across industries, including manufacturing and workforce training. Others encouraged maintaining the conference's interdisciplinary mix while further broadening the diversity of viewpoints represented and devoting additional time to the most thought-provoking policy discussions.

Overall, feedback on the conference was overwhelmingly positive, with respondents describing the event as timely, well organized, and intellectually rich. Participants particularly valued the opportunity to engage with researchers, industry leaders, policymakers, and higher education representatives in a focused setting, with many citing this interdisciplinary mix as one of the conference's greatest strengths. Suggestions for improvement centered primarily on logistics, including allowing additional time for networking, exploring audience-voting tools for panel questions, and balancing the breadth of topics with deeper exploration of individual issues.