OpenAI announced on Thursday that it had banned accounts linked to Russian and Iranian operatives for using its AI tools to support covert influence operations targeting multiple countries. The company disclosed the findings in a safety report, detailing how the accounts employed ChatGPT to generate reports and narratives intended to manipulate public opinion and media coverage.
OpenAI’s enforcement action marks the first time the company has publicly attributed such operations to these actors. The accounts were primarily engaged in two types of campaigns: one focused on placing false narratives in established media outlets and another on spreading disinformation via fake social media accounts. OpenAI’s investigation found that the media-focused campaigns achieved wider reach, with some narratives prompting responses from political figures.
Russian operations targeted Latin American countries, aiming to undermine support for Ukraine and influence local politics, according to OpenAI’s report. One campaign involved operatives impersonating a regional educational authority in Peru to organize school activities about Ukraine featuring Stepan Bandera, a historically controversial figure in Eastern Europe. A Peruvian outlet reported that such an event occurred at a school, and the campaign’s report claimed it had prompted coverage from a Polish news outlet.
OpenAI’s report did not specify the number of accounts banned or the exact timeline of the operations. The company stated that its investigation was ongoing and that it had taken steps to prevent similar misuse of its tools in the future. The findings were published alongside broader discussions about the risks of AI-driven disinformation and the challenges of regulating emerging technologies.
The AI lab emphasized that its tools were not designed for malicious purposes and that the banned accounts violated its usage policies. OpenAI also noted that it had not found evidence of direct coordination between the operatives and their governments, though the campaigns aligned with known geopolitical objectives.
The disclosure comes amid growing scrutiny of AI’s role in spreading misinformation, with policymakers and researchers increasingly calling for safeguards to prevent abuse. OpenAI’s actions reflect a broader trend among tech companies to address vulnerabilities in their platforms while balancing innovation with public safety.