Axios Pro Exclusive Content

Tech unites on deepfake election content

Feb 13, 2024
Illustration of a laptop with binary code and ballot iconography on it

Illustration: Sarah Grillo/Axios

Google, Meta, Microsoft, TikTok, Adobe and OpenAI will pledge to try to mitigate risks around deceptive AI election content, per a draft of an upcoming announcement obtained by Axios.

Why it matters: Tech and AI companies are trying to get ahead of a potential explosion of deepfakes around global elections in 2024, putting their hands up as good actors who are trying to prevent bad outcomes before anyone can accuse them of not being proactive.

Driving the news: Announcement of the pledge is set for Friday at the Munich Security Conference in Germany.

  • News of the draft was first reported in Politico EU.
  • Tech companies have significantly scaled back on allowing political content in past years, but it continues to proliferate online and be seen as an important tool for campaigns.
  • The advent of widely available generative AI tools is invoking worry that the problem of political misinformation online will be worse than ever.

What they're saying: "We will continue to build upon efforts we have collectively and individually deployed over the years to counter risks from the creation and dissemination of Deceptive AI Election Content and its dissemination, including developing technologies, standards, open-source tools, user information features, and more," the draft announcement reads.

  • The companies alone can't find the problem, they concede.
  • "We recognize that no individual solution or combination of solutions, including those described below such as metadata, watermarking, classifiers, or other forms of provenance or detection techniques, can fully mitigate risks related to deceptive AI election content, and that accordingly it behooves all parts of society to help educate the public on these challenges."
  • Adobe said in a statement: "In a critical year for global elections, technology companies are working on an accord to combat the deceptive use of AI targeted at voters."
  • "Adobe, Google, Meta, Microsoft, OpenAI, TikTok and others are working jointly toward progress on this shared objective and we hope to finalize and present details on Friday at the Munich Security Conference."

Details: The voluntary framework is based on seven principles: prevention, provenance, detection, responsive protection, evaluation, public awareness and resilience. The signatories commit through 2024 to:

  • develop technology to mitigate deceptive AI election content
  • address content in a matter consistent with both free speech and safety
  • share best practices with one another
  • update the public on findings
  • offer resources to researchers looking to stem the same risks.

The bottom line: Companies responsible for deploying AI tools to the public know they have to brace for elections or risk disaster.

  • The draft could change by Friday, and more companies may sign on.
Go deeper