Last week was the 1st disclosure from OpenAI confirming that state-linked actors were using their tools for influence operations —something that’s always been a question of when, not if, if it would happen.
But we also got a disclosure from Meta, incl some of the same ops — bc after you create content you have to spread it. Theirs showed that this time, it got no pickup…
Threat intelligence indicators also released by Meta suggest that not all companies are acting to intervene.
So in the above piece for MIT Technology Review, we point out why disclosures matter, and why data sharing is key.
I’m also going to re-up this one from earlier in the week, which was an SIO analysis of threat indicators from late last year - some platforms were slow to act.