OpenAI has introduced a new ChatGPT feature aimed at restricting the generation of realistic images related to polling locations, ballots, and election logistics. This tool, launched on October 11, 2026, is part of a broader election integrity strategy announced by the company in May 2026.
The company is implementing technical safeguards, including the use of Google's SynthID watermarking technology and adherence to C2PA metadata standards. Additionally, OpenAI has partnered with Democracy Works and the Associated Press to provide users with accurate voting information and live election results.
OpenAI has committed to supporting the Protect Elections from Deceptive AI Act and is providing cybersecurity tools to election officials. From 2024 through October 2026, the company has also disrupted foreign influence operations linked to Russia and Iran.
Independent testing conducted by the Brennan Center for Justice in September 2026 revealed that existing AI safeguards could be circumvented to create fake election imagery. It remains unclear whether the new October restrictions effectively address these identified vulnerabilities.
