Emerging partnerships between AI labs and philanthropies seek to measure and prepare for the tech’s impacts on jobs—but the clock is ticking. Also, the final federal rules for Workforce Pell, and essays on how community college grads will staff the AI economy and the latest on noncredit programs and Workforce Pell. (Subscribe here.)

Anthropic and the Gates Foundation

AI faces a crisis-level backlash, fueled in part by fears about threats the technology may pose to millions of jobs. Meanwhile, a consensus has emerged that the U.S. needs to improve its workforce programs and data infrastructure to respond to the challenge.

A strong majority of both business leaders and workers support policies to help people navigate the AI workforce transition, according to a new Milken Institute–Harris Poll. 

For example, both camps back portable benefits that follow workers between jobs, public-private partnerships, requirements for workforce impact assessments, and government-funded community college training programs. Fully 80% of workers say the government needs to start workforce transition programs now, while 83% of all respondents say AI readiness should transcend partisan politics.

However, even some government officials are skeptical that the feds or states are up to the task, at least without substantial help from Big Tech and philanthropies. The Trump administration isn’t spending new money on workforce systems and remains optimistic about the AI transformation, arguing the tech will create more jobs than it eliminates.

Take the data question. Few would argue that public datasets can adequately track links between AI adoption and jobs, the skills employers need, and wages.

Many states have been developing data systems to better connect employment data with information about education and training programs—a broad category of projects that includes talent marketplaces, credential registries, and learning and employment records. And the U.S. Department of Labor recently launched a $15M grant program to help states with these projects.

But many experts privately say they’re worried those tools might not be able to reach a critical mass of people, or be useful to enough employers. Meanwhile, millions of Americans use an LLM every day.

The clock is ticking, says Julia Lane, a prominent economist and data scientist who founded the Coleridge Initiative. Lane has been sounding the alarm that America’s workforce desperately needs a data overhaul to more accurately measure AI’s impacts on jobs and to inform the training needed to help workers.

“There are consequences to inaction,” she says, including anxious and confused young people and education and training providers who are struggling with how to adjust for AI.

Data on Jobs and Education: A national data infrastructure is one area where Big Tech could play a role—by providing the technical chops to help deliver the sort of user experience consumers expect. Likewise, many observers hope AI labs will help pay for the training and reskilling of displaced workers. The Economist raised eyebrows this week by calling for governments to tax tech companies to fund safety nets for workers threatened by AI.

The top AI labs have made some recent moves in this direction. For example, Google recently rolled out investments and research to better understand how AI will change the economy, as well as training to help workers prepare, with promises of more to follow. And OpenAI says some of the $1B its foundation plans to spend this year will go toward practical solutions to challenges arising from AI’s impacts on jobs and the economy.

Likewise, Anthropic last week announced a partnership with the Gates Foundation that includes a focus on the workforce data gap. The company says it will contribute money, Claude credits, and technical support toward:

“Developing portable records of a person’s skills and certifications to carry across schools and jobs; providing trustworthy career guidance for new entrants in the job market and those who are retraining; and creating tools that link data from training programs to employment outcomes in order to measure which economic mobility interventions improve job and wage outcomes.”

One of the most common ways people use Claude is for information about education, with Americans disproportionately using it to search for jobs. The Gates Foundation did benchmarking work on how the tool and other frontier models performed with those tasks, a foundation official says. The results were mixed, with room for improvement on accuracy, tonality, and bias.

The foundation reached out to Anthropic, and the company answered the call. The collaboration has led to the creation of “public goods,” or strategies for getting better results when people tap Claude for information about education and jobs. That could be as simple as an open-source prompt cheat sheet.

Gates and Anthropic will continue to work on helping students with AI-enabled college and career navigation as part of the deepened partnership. The foundation official says Gates also plans to fund middleware players to stitch together public goods for learners, workers, and institutions.

Meanwhile, Anthropic also has partnered with NextLadder Ventures, a $1B initiative from a coalition of foundations and philanthropists that’s focused on economic mobility. The AI company is providing technical expertise and Claude credits on projects like helping people access public benefits and information about jobs.

The Anthropic collaborations are two-way streets, with the company offering help on tools and projects while also working with organizations to improve the information Claude provides to millions of people about jobs, education, and training. A similar dynamic likely is (or will be) part of how Google and OpenAI interact with the field. The amount of money the companies will offer to back related projects remains an open question.

Closing Window: Anthropic is on pace to hit $50B in revenue this year. Its cofounders have called for the federal government to tax the company and other AI labs to help pay for job-training programs.

Several experts praised the company’s partnership with Gates, and the foundation’s investment in AI-powered education and workforce tools.

“I appreciate the focus on meaningful goals—including portable credentials, career guidance, and connecting training data to actual jobs and wages,” says Amanda Winters, executive director of workforce for the U.S. Chamber of Commerce Foundation. “Those are the right challenges to work on.”

Philipp Schmidt, VP of technology innovation for the Axim Collaborative, says higher ed institutions want to be real partners with AI companies. And he says the Gates collaboration is a good first step for Anthropic.

“Hundreds of millions of people are already using commercial AI tools, and there is huge benefit in working together to improve the outcomes, especially in areas like health and education,” he says. “However, it’s just a first step, and what’s needed are stronger commitments to improving outcomes, implementing guardrails, and addressing the gaps in who has access.”

Julia Lane is helping to lead an effort to create trusted, forward-looking information about how AI is changing local economies and labor markets. The Industries of Ideas project has received funding from the National Science Foundation’s Directorate for Technology, Innovation, and Partnerships. It seeks to provide better data to state policymakers and educators. Officials from several states participated in a recent meeting hosted by the group.

More funding will be needed for projects like Industries of Ideas to quickly develop the capacity decisionmakers need. And, as Ellie Bertani from the Gitlab Foundation recently wrote for Work Shift, AI is moving faster than philanthropy.

The Kicker: “I have a real concern that if philanthropic foundations and government agencies don’t step up to the plate in the next few months,” Lane says, “the energy that has been generated will dissipate and the window of opportunity to effect change will be closed.”

AI + Opportunity

Final Rules for Workforce Pell

The U.S. Department of Education this week issued its final rule for Workforce Pell, which allows federal Pell Grants for low-income students to be used for short-term programs aligned with high-demand jobs. The rules hew to the consensus reached during negotiated rule making, but there were two key changes from the draft rules:

  • Registered apprenticeship programs were given additional flexibility to use training provided by employers, unions, or intermediaries for up to 50% of instruction time. In general, programs must be offered by a federal aid–eligible college or university with no more than 25% of instruction coming from a third-party provider.
  • Students who immediately enroll in additional education will be exempted from programs’ required value-added calculation, which is based on learners’ postcompletion earnings. Such students, however, are not exempted from the required 70% job placement rate.

This fact sheet from the Education Department lays out the full eligibility requirements.

States are racing to implement the policy ahead of the July 1 launch date. Governors were given the responsibility to determine which jobs are high demand and what programs align with them, and states also must set up systems to collect and report data on program completion, job placement, earnings, and other required outcomes.

Initially, only a fraction of the short-term programs in each state are expected to qualify, with the Congressional Budget Office estimating that Workforce Pell will support about 100K students a year by 2034. Noncredit programs may have an especially hard time meeting the requirements, as researchers from the Noncredit Research Collaborative write in Work Shift this week. —By Elyse Ashburn

The REal Deal

Open Tabs

Entry-Level Jobs
Employers say AI tools are more likely to increase than reduce their entry-level hiring, finds a survey of 1,500 executives and senior talent leaders conducted by the Strada Institute for the Future of Work. AI also appears to be shifting entry-level work away from administrative tasks toward more complex responsibilities, a finding that echoes previous research. Respondents cited work experience as the most valued indicator of career readiness.

Expanding Apprenticeship
Employers interested in registered apprenticeships too often find that the system is underfunded and poorly aligned with how industry operates, and that programs are still too hard to start, finds a report from Apprenticeships for America that was informed by a business council. The business leaders called for moving beyond bespoke apprenticeship programs—built around rules rooted in the construction trades—to a well-funded and simple-to-use system.

Widening Imbalance
An aging U.S. workforce, a sharp slowdown in immigration, and the rise of AI will lead to a labor force decline of 3.7%, or 5.9M workers, by 2032, finds a report from the Indeed Hiring Lab. AI will be most transformative in knowledge economy jobs, rather than construction, healthcare, and government, which face labor shortages. The challenge of moving workers between industries to close that imbalance will lead to a rise in structural unemployment.

Online Credentials
The work experiences and skills of jobseekers play a substantially larger role in hiring decisions than do degree program modality and institution type, finds research from Ithaka S+R on employers’ perception of online credentials. Candidates with online degrees face a modest penalty. But business, healthcare, education, and other industries aligned with fields of study that have a strong online presence do not penalize candidates with online degrees.

Data Center Jobs
Counties that get a first large data center see their total private employment rise by up to 5% over five to six years, according to research from the Brookings Institution. Construction employment jumps 11% while information sector roles grow 22%. Wages rise by 3-4% for both current workers and new hires. However, employment effects vary widely by facility type, with hyperscale data centers driving greater gains for the information sector.

Job Moves
Nick Moore has stepped down as acting assistant secretary for the U.S. Department of Education’s Office of Career, Technical, and Adult Education, and as deputy assistant secretary at the U.S. Department of Labor. Moore, who previously led the Alabama governor’s Office of Education and Workforce Transformation, in June will start as the Alabama state adult education director at the Alabama Community College System.

Thanks for reading. If this newsletter was forwarded to you, subscribe here. —PF