Duke University Publishes a 45-Page AI Strategy Blueprint

Duke's AI steering committee released a comprehensive institutional plan calling for an Office of AI Strategy, a faculty guidance hub, and a Provost-level committee to guide five-year priorities. The 45-page report also calls for targeted faculty hiring, expanded student research opportunities, and shared computing infrastructure. Duke's Chronicle and EdScoop both covered the release as a major governance milestone for higher education.

Why it matters for campuses
This is the kind of structured blueprint that separates institutions ready for the AI era from those still reacting to it. Departments can use Duke's framework as a starting point for their own planning conversations.

Read the Duke report

Colleges Are Adopting AI Faster Than They Can Govern It

A new eCampus News analysis found that universities are embedding AI into admissions, curriculum, advising, and enterprise contracts faster than they are building oversight structures. Faculty at California State University pushed to not renew a $17 million OpenAI contract. The University of Colorado delayed student access to ChatGPT until August after faculty raised governance concerns. The piece argues that AI is becoming an organizing logic for institutions before the rules are in place.

Why it matters for campuses
The governance gap is real and growing. Faculty are no longer just asking how to manage AI tools. They are asking who gets to decide whether and how AI belongs in their classrooms and hiring pipelines.

Read the full analysis

Anthropic Accidentally Leaked Details of a New Model Called Mythos

On March 27, Anthropic accidentally revealed the existence of a model called Mythos through a data leak. Anthropic confirmed it is testing the model and described it as a step change in capabilities above Claude Opus 4.6. The model scores dramatically higher on coding, academic reasoning, and cybersecurity benchmarks. Anthropic is holding back the release due to concerns about its power and cost to serve.

Why it matters for campuses
Educators who use Claude today should know a more powerful version is coming. Institutions planning AI policies now should build in flexibility for tools that will look very different in six months.

Read about Claude Mythos

OpenAI Winds Down Sora to Focus on Its Next Major Model

OpenAI announced it is shutting down its Sora video generator to free up computing resources for its next major model, internally called Spud. CEO Sam Altman said Spud is ready in weeks and is designed to accelerate economic productivity. The Sora team will shift to world simulation research for robotics. OpenAI's partnership and $1 billion investment with Disney tied to Sora is also reported to be on hold.

Why it matters for campuses
This shows how quickly priorities shift at major AI labs. Campus technology agreements with AI vendors should include flexibility clauses. Tools that seem permanent today can be discontinued with little notice.