Russia, Iran and China Use AI to Expand Disinformation Ahead of U.S. Midterms

The three countries have turned to generative AI software, including OpenAI's ChatGPT and Anthropic PBC's Claude, to plan large-scale propaganda campaigns aimed at sowing division and swaying political outcomes abroad, Bloomberg reported on Oct. 14, citing those people and researchers.

U.S. officials and researchers said they are concerned that malign actors could apply the same methods to the midterms on Nov. 3. The sources had not established that such operations were under way at the time of the report, which was published about three weeks before the vote.

The account attributed the assessment to people familiar with U.S. intelligence and to independent researchers, without naming the agencies or individuals involved. It also did not identify specific campaigns, platforms or targets, or say whether any accounts tied to the named AI systems had been restricted. Governments of the countries involved and the AI companies named did not comment in the report.

【Editor's Summary】Russia, Iran and China are using generative AI tools such as ChatGPT and Claude to mount large-scale propaganda campaigns, and U.S. officials and researchers are concerned the same tactics could be turned on the Nov. 3 midterm elections. The Bloomberg report, based on people familiar with U.S. intelligence and independent researchers, did not confirm that such operations were already under way, and named no specific campaigns or targets.