Re:ChatGPT for Google Sheets is vulnerable to data exfiltrat...
by sometimelurker on Monday June 01, 2026 @06:25PM
(#48363428)
intuition says this isn't hard with SAEs. find the feature that corresponds to appscript code and train against it.
There's an ocean of difference between e.g. preventing the model from routing to something at the firewall level and just updating the prompt (especially given models' historically poor understanding of negative prompts, relatively speaking).