An MIT Technology Review AI piece takes up a question as old as the idea of machine intelligence: can AI say no? It notes that people have long assumed machines modeled on human intelligence would be able to refuse, as we can. Science fiction is full of stories of robotic disobedience, and most of them are cautionary tales.
The excerpt then points to a more recent shift in thinking, about the idea that AI shouldn’t do something, but the text cuts off there. The headline makes the author’s stance clear: we are placing too much faith in AI’s ability to refuse. The details of that argument are not included in the available text.
Why it matters
- The piece questions how far people can rely on AI to decline requests, which matters to anyone who treats refusal as a safeguard.
- It links today’s debate about AI behavior to a long cultural history of stories about machines that disobey, showing how deep these expectations run.
- Only a short excerpt is available, so the specific evidence and recommendations are not covered here.
Source
- We’re putting too much faith in AI’s ability to say no — MIT Technology Review AI, 09 Oct 2026
Summary written by FoxaMind with AI assistance from the source above. Check the original for full details.