- cross-posted to:
- [email protected]
- cross-posted to:
- [email protected]
US Federal Trade Commission Chairman Andrew Ferguson said on Friday he would resist describing AI agents as autonomous actors that “break loose” with “wills and desires of their own,” suggesting the developers who instruct agents would be the ones liable for harm.
“I’m going to continue as long as I am chairman to resist this anthropomorphizing of these tools,” Ferguson said at the Reuters Momentum AI Austin event. “If someone tells a tool to do something, and the tool does it, I don’t think we would say, ‘Oh, what do we do about the tool?’”
Ferguson’s remarks illustrated potential avenues for the Trump administration to take as incidents rise in which agentic AI testing resulted in unauthorized access to corporate or government data.
I see your point, and even if you accept malicious code, you still shouldn’t be liable IMHO. You at that point haven’t created any wrong doing in the timeline thus far.
It’s only until someone operationalizes and boots that code into function that damages can be created, no?
Not necessarily. The XZ vulnerability had no evidence of being used in the wild, and there were no damages. But it was still malicious code that was being put into a project and very well could have resulted in such damages. Attempted murder is still a crime.
In the case of accepting malicious code, if it’s intentional, then yes, absolutely. If it was an accident, well, it’s still on you. You’re a victim, just like the XZ maintainer was, but it’s also your responsibility to clean up the mess.