Skip to main content
Close-up of a smartphone displaying an AI-generated deepfake of Sam Altman, highlighting the rise of AI deepfakes on Instagra

Editorial illustration for Not Even Sam Altman Is Safe From AI Deepfakes on Instagram Now

AI Deepfakes of Sam Altman Spread on Instagram

• 4 min read

Alexandr Wang, Meta's chief AI officer, posted a short AI-generated clip earlier this month showing himself, Mark Zuckerberg, and OpenAI CEO Sam Altman on a fake four-way video call with Jolly, the mascot for Meta's new consumer AI agent, Muse. Altman's digital double shows up dressed as Jolly, furry hood and all, with face paint done up to match the character. The caption reads "When the homies meet about openai product strategy," a jab at the AI agent race Muse just entered.

None of this breaks any rules. Meta's own AI policy allows employees to post this kind of content, even when it depicts a rival company's CEO without his involvement. That's a shift from a few years back, when deepfakes of real people, politicians especially, drew public condemnation and occasional legal threats. During this year's midterms, political groups spent millions on AI-generated attack ads, and nobody blinked much.

Wang has leaned into what some have called "anti-PR" for Muse, posting constantly on social media since launch. The Altman clip is the latest example of how casually AI depictions of real executives have slid into normal marketing.

Earlier in November, Alexandr Wang, Meta’s chief AI officer, shared a short clip to promote Muse, the company’s consumer AI agent. In the AI-generated clip, Wang, Meta CEO Mark Zuckerberg, OpenAI CEO Sam Altman, and Jolly, the anthropomorphic mascot of Muse, are on a four-way video call. When Altman joins the call he’s dressed like Jolly, wearing a furry hooded suit and face paint to look like the cutesy character.

Why this matters

Alexandr Wang running an OpenAI deepfake through Meta's own policy exceptions tells us something about where the rules actually stand. Meta built carve-outs for "parody" and "satire" into its AI labeling requirements, and its own chief AI officer is now the test case for how far those carve-outs stretch. That's a conflict of interest dressed up as content moderation.

For founders building on Meta's platforms, this is worth watching closely: if the company's top AI executive can post a fabricated video of a competitor's CEO and call it compliant, the line between "allowed synthetic media" and "targeted disinformation" is whatever Meta decides it is that week. Three years ago, a video like this would have triggered apology statements and policy reviews. Now it's a marketing post.

Researchers tracking platform trust should note the midterm-election deepfakes referenced here aren't a separate problem from Wang's video, they're the same infrastructure. The question isn't whether Meta can detect synthetic media. It's whether it wants to, when the beneficiary sits in the C-suite.

Common Questions Answered

What AI-generated deepfake did Alexandr Wang post on Instagram featuring Sam Altman?

Alexandr Wang, Meta's chief AI officer, posted an AI-generated video clip showing Sam Altman, Mark Zuckerberg, himself, and Jolly (Meta's Muse AI agent mascot) on a fake four-way video call. In the deepfake, Altman appears dressed as Jolly with a furry hood and matching face paint, with the caption joking about OpenAI product strategy discussions.

What is Muse and why did Meta's chief AI officer create this deepfake?

Muse is Meta's new consumer AI agent, and Alexandr Wang created the deepfake video as a promotional tool to showcase the technology. The video served as a playful jab at the competitive AI agent race that Muse had just entered, featuring the company's mascot Jolly alongside tech industry leaders.

How does Meta's policy on AI-generated content relate to this deepfake incident?

Meta has built carve-outs for 'parody' and 'satire' into its AI labeling requirements, which Wang's deepfake appears to fall under. However, this raises concerns about conflicts of interest, as Meta's own chief AI officer is testing the limits of these policy exceptions, potentially setting a precedent for how far these carve-outs can stretch.

Why is this deepfake incident significant for developers building on Meta's platforms?

The incident highlights a potential conflict of interest in Meta's content moderation policies, as the company's top AI executive is using policy exceptions to create deepfakes of other tech leaders. This raises questions about how consistently and fairly Meta will enforce its AI content rules for other platform developers and creators.

LIVE05:12California Takes OpenAI to Court in New Lawsuit Claiming Private Ownership