OpenAI is undertaking a series of new efforts to combat AI-generated misinformation ahead of upcoming elections in the US and other countries.
The ChatGPT-maker is working with a US-based tech non-profit called Democracy Works to display reliable information about voting and registration processes, including voting locations and other election logistics, when people ask about those topics, according to a blog post published on Wednesday, May 27.
It is also teaming up with The Associated Press to show users live counts during upcoming elections in the US and Brazil. “Globally, we will continue to refine the way web search surfaces helpful information with source links,” the company said.
In the past few years, a growing number of election candidates have turned to AI tools as part of their campaigns while voters continue to seek election-related information from AI chatbots. This indicates that AI is fast becoming an election mainstay.
However, there are rising concerns about the potential use of these AI tools to influence election results. Similar to social media platforms, AI companies such as OpenAI are coming under increasing scrutiny for inaccurate and misleading information being generated by chatbots. With the rapid advances of the technology in 2026, voters are set to face a new wave of AI-driven disruption.
Meanwhile, OpenAI is offering US state election authorities access to its cybersecurity products such as Codex Security. Registered voting system manufacturers in the US are also being folded into OpenAI’s Trusted Access for Cyber programme. The Sam Altman-led company said it has also briefed the US National Association of Secretaries of State and the US National Association of State Election Directors on the latest cyber capabilities of its AI model.
Last month, Anthropic announced Claude Mythos, its most powerful AI model to date, is said to be exceptional at scanning software systems and identifying previously unknown (zero-day) vulnerabilities in limited timeframes. A week later, OpenAI unveiled a similar model called GPT‑5.4‑Cyber.
Both companies have released their respective models to a limited group of partners for testing. These developments have sent governments and organisations around the world scrambling to prepare for how attackers might eventually use such models to carry out cyber attacks. More recently, OpenAI introduced a new AI cybersecurity platform called Daybreak that looks to help companies find software weaknesses, detect cyber threats, and fix security issues faster using AI tools.
“To help combat misleading “deepfakes”, we are investing in a multi-layered provenance approach that will equip people to verify whether content they’re seeing has been created or modified with AI,” OpenAI said.
Last week, OpenAI announced a new partnership with Google to include the invisible watermark called SynthID on all visuals generated using its AI products. Additionally, OpenAI launched a new public verification tool which allows users to easily find out whether an image was generated using AI or not.
Stating that provenance tools are not a fool-proof solution to tackle election-related deception, OpenAI said it is “happy to partner with social media companies as they consider steps to protect elections by using provenance markers as a relevant signal in deciding which civic content they recommend and distribute in people’s feed.”
OpenAI said that its current usage policies prohibit the use of its AI products to create or distribute scaled campaign messaging for or against a candidate, political party, or ballot measure.
However, political campaigns are allowed to use OpenAI’s products for human-directed work, like drafting internal briefings, planning, everyday writing, translation work, compliance, and administrative tasks. “Consistent with our rules against using our services for scaled campaign advocacy, we will not allow advertisers to run political ads on our platform this cycle,” OpenAI said.
In its announcement post, OpenAI endorsed legislation aimed at combatting AI-generated deepfakes and promoting transparency such as the Protect Elections from Deceptive AI Act and Preparing Election Administrators for AI Act.
For context, would add new prohibitions to knowingly distributing materially deceptive AI-generated content involving federal candidates when used for federal election activity. The Preparing Election Administrators for AI Act would require the Election Assistance Commission to issue a report with voluntary guidelines for election offices on the use and risks of AI technologies in election administration.
“2026 is the world’s second major election year since generative AI became widely available, and we’re continuing to build on the foundation we laid in 2024 to help protect elections in countries and territories around the world,” OpenAI said.
“Our focus is to build and responsibly deploy groundbreaking products in ways that: surface reliable information about voting and results, support cyber infrastructure defenders, increase transparency around AI-generated content, combat misuse by bad actors, and monitor bias in our models to keep ChatGPT’s responses politically neutral,” it added.
