AI-Assisted Influence Operations Become Routine
State-linked information operations keep industrializing content production with commercial language models and adapting specifically to defeat AI-detection heuristics, so provider takedown reports become a recurring and contested source of attribution that platforms and regulators cannot independently verify.
weakening · confidence 36 · Emerging (watchlist) · tracking since August 26, 2026 · updated August 28, 2026
Score history
Daily conviction score, 0 to 100. Higher means the thesis is more strongly corroborated.
Now 36 · -2 since Aug 27 · ranged 36 to 38
Showing the last few days. Unlock full score history.
Why the conviction moved
- Aug 26Strengthened +6
OpenAI banned Russia-origin accounts that used VPNs to reach ChatGPT from a blocked country to run a fake Israeli think tank and a pro-Russia 'Sovereignty Index,' explicitly instructing the model to strip machine-written tells, and rated the campaign category three of six on Brookings' Breakout Scale. Prompting specifically to defeat AI-detection heuristics is the adaptation step the thesis predicts, and the provider's own scale rating is again the only attribution outsiders can check.
Source trail
Supporting · August 26, 2026
OpenAI Bans Russia-Origin Accounts Running a Fake Israeli Think Tank and a Pro-Russia 'Sovereignty Index'
OpenAI banned Russia-origin accounts that used VPNs to reach ChatGPT from a blocked country to run a fake Israeli think tank and a pro-Russia 'Sovereignty Index,' explicitly instructing the model to strip machine-written tells, and rated the campaign category three of six on Brookings' Breakout Scale. Prompting specifically to defeat AI-detection heuristics is the adaptation step the thesis predicts, and the provider's own scale rating is again the only attribution outsiders can check.
OpenAI
Unlock full source trail, score history, and daily updates.
Unlock TrendsAffected regions & assets
Townsquare
Argue the thesis in Townsquare.