AI is eroding office hours, study groups, and the trust between faculty and students, MIT report finds

What happened

An expert committee at the Massachusetts Institute of Technology has published a report warning that generative AI is hollowing out core elements of the undergraduate experience, including office hours, study groups and the institution's flagship undergraduate research programme. The committee finds that trust between faculty and students is deteriorating as AI tools reshape how coursework, research and mentorship are conducted on campus.

According to the report, some faculty members are now weighing the use of AI agents as substitutes for student research assistants, a shift that strikes at one of the most formative relationships in academic life. The committee argues that these pressures amount to more than isolated classroom disruption and instead call for a wholesale rethink of how higher education is structured and delivered.

The intervention carries particular weight given MIT's standing as one of the institutions most closely associated with the development of modern AI, lending the warning added authority within the sector.

Why it matters

This is fundamentally a story about what happens when a powerful new technology is adopted faster than the human systems and norms built around it can adapt. Office hours, study groups and mentored research are not incidental features of a university; they are the service design through which trust, feedback and accountability are built between students and faculty. When AI tools substitute for those interactions rather than supporting them, the underlying relationship — not just the task — starts to erode.

For leaders in experience and digital transformation well beyond education, the lesson generalises: AI adoption that optimises for individual efficiency can quietly undermine the relational infrastructure — trust, mentorship, shared accountability — that organisations rely on to function well. A report from an institution at the centre of AI's development calling for structural reform is a signal that this tension is now serious enough to require deliberate redesign, not just policy tweaks.

The Renascence take

The headline risk here isn't that students are using AI — it's that institutions are letting AI fill gaps in trust rather than asking why those gaps exist and redesigning around them deliberately.

Most organisations treat AI erosion of trust as a compliance problem to police rather than a service-design signal to act on. If students are outsourcing to AI instead of attending office hours, the deeper issue is often that the human interaction wasn't delivering enough value to compete with a faster alternative. The fix isn't banning the tool — it's redesigning the human touchpoints so they offer something AI genuinely cannot: judgement, relationship and accountability. Any customer-facing or employee-facing organisation watching AI adoption accelerate should ask the same question MIT is now asking of itself — not "how do we stop this," but "what were these interactions actually for, and are we still delivering it?"