Tech giants make commitment against political deepfakes

On February 16, 2024, during the Munich security forum, around twenty major digital corporations signed a declaration that aims to combat political deepfakes. The document outlines overarching principles and means of action to “help prevent deceptive AI-generated content from interfering in elections planned this year around the world.”

Among the signing parties are Adobe, Amazon, Google, Meta, Microsoft, OpenAI, Snapchat, TikTok and X. The statement addresses “convincing audio, video and photo content generated by AI that deceptively fakes or alters the appearance, voice, or actions of political candidates, election officials, and other key players in a democratic election, or that provides false information to voters about when, where, and how they can lawfully vote.”

The parties pledge to develop common tools to identify and track content generated by AI, in line with the C2PA tracing standard, which is currently being established. The digital giants also want to improve the ability of large platforms to identify deceptive content. To this end, they suggest using detection software, features that allow creators to make AI-use disclaimers, and reporting tools.

The document is much more vague in regard to the measures to implement in the event suspicious content is detected. The signing parties must indeed strike a balance between the fight against deceptive content and freedom of speech. Moreover the declaration specifies that they will have to “pay attention to context, and preserve educational, documentary, artistic, satirical as well as political expression.”