AI
Meta's AI Chief Posts Deepfake of OpenAI CEO on Instagram—and It's Totally Allowed
As AI-generated content becomes routine on social platforms, Meta's own executives are using the technology to mock competitors, raising questions about oversight and consent.
1/2
The Verge
Instagram's feed increasingly blurs the line between real and synthetic media. The latest example involves Meta's chief AI officer creating and sharing an AI-generated video of a prominent rival—a move that highlights how casual tech executives have become with deepfakes, even as the technology raises ongoing concerns about consent and misuse.
The Video and the Context#
Alexandr Wang, Meta's AI chief, posted a short synthetic clip promoting Muse, the company's consumer AI agent. The video depicts a video call between Wang, Meta CEO Mark Zuckerberg, OpenAI CEO Sam Altman, and Jolly—Muse's animated mascot. In the fabricated scene, Altman arrives dressed as Jolly in a furry suit and face paint. The caption reads "When the homies meet about openai product strategy," a jab at the growing AI agent competition and Muse's market positioning.
The post comes as Wang has pursued what observers have called an aggressive promotional style, tweeting hundreds of times and dismissing traditional marketing as "corposlops." For Wang and Meta, using AI to lampoon a competitor represents both a marketing tactic and a demonstration of the technology itself.
What Meta's Rules Allow#
Meta's current AI policy requires synthetic videos depicting real people to be labeled as such—which Wang's post carries. However, static images face no disclosure requirement. This gap became evident after an AI-generated image of actress Mia Goth circulated on Instagram in October; because it was a still image, no AI label was required. Her representative told The Verge the situation was "incomprehensible," given that the company had recently faced criticism over sexualized deepfakes of public figures.
The inconsistency matters because verification badges—even when issued to fan accounts—can amplify reach and lend false credibility to synthetic content.
A Shift in Norms#
Just years ago, deepfakes depicting real people triggered industry-wide backlash and legal threats. In 2024, Scarlett Johansson pursued action against OpenAI over voice similarity concerns. Today, AI-generated content has become woven into the social media landscape. Political campaigns ran millions of dollars in AI ads during recent midterm elections. When OpenAI launched its Sora video app, the platform was with fabricated Altman clips.

