Never in recent memory have colleges, universities and their endowments been under more scrutiny—and more pressure to perform in less-than-ideal conditions.
Some of the wealthiest colleges and research universities will feel the sting of a large tax increase on endowments, including imposition of an 8% rate on a handful of institutions and a 4% rate on others.
For schools subject to the higher rates, this could mean allocating more dollars to taxes and fewer to financial aid, research programs, student services, faculty and other critical areas.
Further complicating matters for institutions and their endowments are growing regulatory and compliance requirements, stricter rules governing self-dealing and executive compensation, heightened uncertainty about nonprofit tax exemptions, and potential reductions in federal research grant funds.
What’s more, donations to endowments are down. New gifts to the more than 650 institutions participating in a National Association of College and University Business Officers (NACUBO) and Commonfund study declined 9.2% in fiscal 2025.
All these factors put added pressure on endowment spending, which has increased an estimated 17% in the past two fiscal years, according to the NACUBO study.
That underscores “how important well-managed endowments are to colleges and universities,” said NACUBO’s Kara D. Freeman. “Endowments help fuel innovation and serve as a stable foundation for institutions.”
AI advantages for endowments
How can institutions preserve the stability of their endowments in the face of so many destabilizing forces? Leaning more heavily on intelligent technologies and data certainly can help.
A range of promising use cases involving AI agents, generative AI and intelligent automations have become viable. Here are several to consider:
• Risk management. The battery of challenges confronting institutions and their endowments puts a premium on capabilities that can analyze various potential outcomes across a range of risks and stressed-revenue scenarios.
AI agents can perform modeling and stress-test tasks, then report back with their findings. They also can execute regular risk reviews inside an endowment’s investment portfolio to gauge how closely the portfolio tracks to the endowment’s core principles and objectives (ensuring adequate liquidity, meeting spending and preservation benchmarks, etc.) and flag risks or anomalies that need addressing.
Doing all this in advance helps investment decision-makers stay disciplined when their endowments do indeed come under stress.
• Investment analysis and due diligence. Oxford University Endowment Management, which oversees a fund worth roughly £6 billion (about US$8 billion), is using generative AI to review and analyze a vast number of lengthy, unstructured, non-public investor reports from equity funds.
A natural language interface digests the reports and provides insights to analysts, who then can interact with the tool conversationally and verify the accuracy of its findings via citations. The result: faster analysis and decision-making, so investment staff can focus more on the fiduciary judgment required to meet endowment goals.
It’s also a critical step toward turning endowment management into an autonomous enterprise.
• Audit readiness. Here’s another area where AI is proving invaluable, helping investment officers by tagging every action, and creating and maintaining audit trails, documentation of provenance, and tax reporting documentation.
AI agents also can identify a broken audit trail, flag and find missing documentation and source attributions, and ensure data is properly formatted and presented so it tracks to the requirements of a specific type of audit.
• Compliance. AI agents can keep teams abreast of the latest policy and regulatory changes and ensure an endowment’s actions and reporting comply with them. That’s especially helpful in light of the tax policy changes impacting endowments in 2026, such as new excise taxes on executive compensation.
• Evaluating educational program investments. Stretched endowment dollars put a premium on program investments that yield a high ROI. Using AI tools, institutions can model and assess the likely impact (on enrollment, revenue, etc.) of an investment in a new program of study, for example.
• Ensuring gifts are used as directed. Using AI agents, institutions can monitor restricted gifts to ensure funds are used in accordance with donor terms, and create an audit trail to document to donors, internal auditors, and regulators how they were used.
Bringing these use cases to life starts with high-quality data: contextual, semantically rich, current, comprehensive and trustworthy. AI can deliver insight and execute tasks only when it is grounded in enterprise-wide business data, connected to systems of record, and governed through transparent, secure controls.
Where is all this leading?
We’re headed toward a more autonomous model for endowment management, where AI agents are grounded in trusted institutional data, connected to end-to-end financial and compliance processes, and governed with the transparency, security, and oversight required for mission-critical decisions.
In practical terms, that means routine analysis, documentation, monitoring, and workflow coordination can increasingly be handled by intelligent systems, while people remain firmly in control of strategy, fiduciary judgment, donor stewardship, and accountability.
The benefit is not autonomy for its own sake, but a more adaptive, resilient enterprise that can move from insight to action faster, preserve resources, reduce risk, and keep endowments focused on fulfilling their long-term mission even in turbulent times.