In 2019, the University of Michigan invested $20 million in OpenAI. The position was listed in annual reports under "alternative assets." Nobody told students or faculty.
That stake is now worth $2 billion. It was only recently discovered in court documents after Elon Musk sued OpenAI.
Michigan is not an outlier. Our analysis of 90 leading American universities revealed an alarming pattern: institutional endowments are quietly funneling capital into the very AI companies eliminating the jobs students are paying to pursue — often with no disclosure, no press release, and no obligation to ever say a word. Billions more from corporate partnerships, gifts, and government contracts are being poured into AI ventures.
You can view the full results of our analysis, and your university's report card, here.
The Endowment Scandal
University endowments don't usually invest in AI companies directly. They deploy capital as limited partners in VC funds (such as Sequoia Capital, Andreessen Horowitz, Thrive Capital, Khosla Ventures) who hold positions in OpenAI, Anthropic, xAI, and dozens of other AI companies.
At Michigan, documents obtained by Fortune confirm LP positions in Sequoia, Andreessen Horowitz, Accel, General Catalyst, and Y Combinator. These five firms have collectively deployed more capital into AI than any other institutions on Earth. In a single board vote in September 2022, Michigan committed $75 million across seven Sequoia funds. On top of the $2 billion OpenAI position. On top of a separate commitment to Andreessen Horowitz. On top of an additional $180 million to Sam Altman's personal AI venture fund, Hydrazine Capital.
This is one school. There are 89 more.
The University of California Regents oversee approximately $190 billion across ten campuses including Berkeley, UCLA, and UC San Diego. State law requires the regents to publish their private equity holdings, revealing significant AI exposure: over $1.4 billion committed to Sequoia and its affiliate funds, $157 million to Khosla Ventures (which holds direct OpenAI equity), and $235 million to GGV/Notable Capital. Niche AI firms such as The Engine Fund, NeoTribe Ventures, The House Fund, and Vertical Venture Partners account for an additional $205 million. An exact total is impossible to calculate because UC is not required to release itemized portfolio holdings, and numerous other funds include additional AI exposure. While this information is technically public, it is buried in dense financial reports, leaving UC's 299,000 students completely unaware.
The University of Texas system is one of the only other institutions whose holdings are partially visible. UTIMCO manages approximately $44 billion across UT and Texas A&M and has disclosed venture positions in Thrive Capital, Sequoia, and GGV/Notable Capital. Thrive's 2022 fund posted a 126% internal rate of return as of late 2025, the highest single-vintage IRR ever recorded in UTIMCO's public disclosures. It is anchored by early OpenAI and Cursor positions. UTIMCO's own CIO publicly acknowledged the fund is currently 5% overweight in AI exposure.
Most public institutions, however, have no disclosure obligation. The University of Virginia manages $15.5 billion through UVIMCO and allocates 58% to alternatives including venture capital. In their FAQ, they openly state it is "contractually required to preserve the confidentiality" of its fund managers. In simpler terms: a public university is using private contractual secrecy clauses to hide where institutional money goes.
The investments of private universities, on the other hand, are notoriously opaque: Not one currently discloses endowment holdings. Despite being kept secret, an independent analysis from Markov Processes International has managed to shed light on two elite schools. MPI reports that in 2025, a standard alternatives-heavy endowment portfolio should have returned 11–12%. Yet, MIT returned 14.8% and Stanford returned 14.3%. MPI suggests this gap is best explained by concentrated exposure to AI and digital assets. Neither university has challenged this analysis, nor provided public transparency on their specific allocations.
A Story Bigger
Than Endowments
Our full analysis revealed a series of controversial AI entanglements beyond endowments, totaling a figure in the billions. The University of Michigan, for instance, is financing $850 million of a $1.2 billion AI computing center for nuclear weapons development and private-sector research. The university purchased land and informed the local township supervisor about their plans by text message. After fierce resistance from the local community, Michigan is now eyeing a low-income area already in the 85th percentile for pollution exposure to house the facility.
At UC Berkeley, Meta, Amazon, Google, and Microsoft directly co-fund and co-direct AI research inside Berkeley's own labs through the BAIR Open Research Commons. Corporate researchers are embedded directly within university labs, co-authoring papers and blurring the line between academic independence and commercial interests. Faculty members simultaneously hold paid positions at the same companies. An additional $12 million fund was launched in 2025 to convert that university research into private equity.
At the University of Washington, technology companies such as Amazon have effectively annexed the Paul G. Allen School of Computer Science, with millions in corporate research funding supporting academic initiatives. Faculty members hold simultaneous Amazon Scholar appointments, and research is directed at Amazon's commercial robotics priorities. Internal Amazon documents, obtained by the New York Times, project the replacement of more than 600,000 jobs with autonomous technology by 2033. The university research program and Amazon's displacement plan are not separate. They are two sides of the same project.
These examples are merely a shortlist, and do not touch upon several university-backed datacenters, a plethora of corporate partnerships, gifts, and military contracts funding AI for war in dozens of universities nationwide. The full school-by-school breakdown is available here.
The Students
Paying For It
The class of 2025 entered one of the worst white-collar job markets in a generation. Overall unemployment for recent graduates reached 5.8% in early 2025, though that understates the damage: underemployment, or graduates working jobs that don't require a degree, reached 41.3%. Furthermore, entry-level postings in finance, law, consulting, and media fell by double digits in 2024 and 2025 while enterprise AI adoption simultaneously accelerated.
The Burning Glass Institute found that the gravest declines were concentrated in precisely the knowledge-work categories most new graduates enter: data analysis, paralegal work, content creation, basic software development, and financial modeling.
The evidence continues: JPMorgan told managers to maintain a strong bias against new hires as it deploys AI. Goldman's CEO announced the bank would reorganize its people around AI productivity. A Stanford working paper found employment for 22-to-25-year-olds in AI-exposed occupations (defined as software, accounting, clerical work, and customer service) dropped 13% since 2022. Goldman Sachs estimated in April 2026 that AI is eliminating approximately 16,000 American jobs per month, with Gen Z absorbing a disproportionate share (a figure that likely undercounts the true impact because it doesn't capture hiring slowdowns that compound displacement, as noted by Goldman's own economists).
These findings are not projections. The culling of young professionals has already begun — and the universities funding AI technology know this. Their own researchers publish the papers documenting it. They warn about it at commencement ceremonies. And then, they cash in.
The Corporate
Seizure of
Higher Education
Legislation requiring universities to disclose itemized endowment holdings has been introduced in Congress multiple times. Universities spent $37 million lobbying Congress in 2025. Endowment policy was a top priority. No legislation passed.
That is, perhaps, the clearest possible statement of whose interests higher education institutions now serve. American universities were built on a paramount, foundational premise: that knowledge belongs to the public, and that institutions entrusted with educating a generation are accountable to that generation.
Today, that premise has been discarded in favor of predatory investing, corporate collaborations, and other backroom deals with technology companies and billionaires. Instead of an education that provides social mobility, students receive a credential for a job market being systematically dismantled by the same institutions their tuition keeps alive.
And the universities won't stop until it costs them something.
Right now, it costs them nothing.
This analysis does not constitute financial or legal advice.