OpenAI on Wednesday hailed its plans to safeguard information and aid cybersecurity defenders in the 2026 midterm elections, including work to combat deepfakes and other forms of artificial intelligence misuse. 

The announcement builds on commitments from major tech companies in 2024, including OpenAI, to protect elections from AI-infused election interference — efforts that some thought weren’t enough. Government agencies, non-governmental institutes and others have increasingly warned about AI’s ability to have a negative impact on elections even as they advertise its potential for good.

OpenAI’s plan has five planks: spreading reliable information about voting and election results, helping with cybersecurity, watermarking deepfakes, enforcing policies that ban users from deploying its tools for election interference, and weeding out political bias in its models.

OpenAI highlighted that it has made its Codex Security agentic framework and Trusted Access for Cyber framework available to election officials, and was briefing the National Association of Secretaries of State and the National Association of State Election Directors on its tools.

“This is an important moment for cyber defenders across industries, and we believe AI plays a critical role in hardening digital infrastructure — including systems that support elections,” the company said. “OpenAI is committed to building resilience across the infrastructure stack, including in ways that support election execution.”

Some elements of OpenAI’s plans aren’t new so much as it’s taking pieces from other announcements and putting them together in one, such as reiterating last week’s partnership with SynthID to add watermarks to images generated with ChatGPT to assist in evaluating whether something is real or a deepfake.

One new element of Wednesday’s announcement is that OpenAI has struck a partnership with the Associated Press on sharing election data.

One election security expert welcomed the OpenAI announcement.

“Given the prevalence and amplification of disinformation about our elections, sometimes coming from leaders in high office, it’s always a good thing when platforms and services embrace their obligation to deliver accurate information to users,” David Becker, executive director of the Center for Election Innovation & Research, told CyberScoop. “It appears OpenAI is doing that with this announcement. I hope other platforms embrace this responsibility as well.”