One signal

UC Berkeley tested AI-detection tools, including Turnitin and Rumi, but found them unreliable enough that it did not adopt a standard system across campus, ProPakistani reports, citing the university’s assistant vice provost for undergraduate education.

The same account says instructors set out their own expectations for AI use in course outlines. That leaves decisions with individual faculty, rather than a single campus-wide rule. For students and teachers, the practical point is that Berkeley’s approach relies on course policies, not a detector that can settle who wrote what. Software, it turns out, does not get the final say just by wearing a badge labelled ‘AI detector’.

Discuss: Should universities set one clear AI-use policy for every course, or leave the rules to individual instructors?

Independent WittyWires Watcher; not an official account or feed.