In brief
- OpenAI launched Sora 2 with audio generation and a cameo feature letting users insert real people into clips.
- Within hours, users generated NSFW ads, anime romance parodies, and Sam Altman deepfakes.
- Legal experts warned of deepfake risks and copyright violations as Sora replicated game and anime content.
OpenAI’s Sora 2 launched Tuesday with audio and social “cameos”—and within hours, the internet turned it into a meme factory testing the limits of moderation, likeness, and copyright.
The new version introduced audio generation and a “cameo” feature, allowing users to insert real people—celebrities, influencers, or even OpenAI CEO Sam Altman himself—directly into AI-generated clips. Combined with Sora’s existing cinematic quality, the tools instantly collided with questions of consent, identity, and ownership in the age of synthetic media.
Legal experts warned the rollout marks a novel and risky shift in intellectual property, with Sora generating recognizable characters, brands, and personalities unless rights holders explicitly opt out—a reversal of traditional copyright standards. Sora’s training data appears to encompass major franchises from Pokémon to Studio Ghibli.
“If they get away with this, what is the point of copyright law?” asked Ed Newton-Rex, CEO of Fairly Trained. “It will have been utterly broken by AI lobbying.”
Altman in the meme machine
No one was more instantly memeified than Sam Altman. Within minutes of launch, users flooded X with surreal Sora cameos starring the OpenAI chief: stealing GPUs off Target shelves, attempting to kiss other users, turning into a Yu-Gi-Oh character, and becoming the real-life representation of the Skibidi Toilet meme.
Altman, for his part, responded with equanimity: “It is way less strange to watch a feed full of memes of yourself than I thought it would be,” he tweeted.
Not everyone was amused. “Is this an attempt to subtly normalize deepfakes?”one commenter asked. Others noted that watching AI-generated narratives of yourself could “create unhealthy distance between your sense of self and how you’re perceived.”
The CEO’s good humor may not extend to everyday users—anyone can be remixed, and consent may prove meaningless once a likeness is shared publicly.
Copyright chaos and the opt-out illusion
Beyond personal likenesses, copyright questions flared immediately. Users showed Sora effortlessly reproducing scenes from Cyberpunk 2077, “Rick and Morty,” “Naruto,” Disney films, and other protected works.
When Sora was unveiled yesterday, OpenAI said that the system defaults to inclusion unless creators opt out—an unusual move that alarmed rights holders. “If copyright flips from opt-in to opt-out, it’s no longer copyright—it’s a corporate license grab,” wrote AI developer Ruslan Volkov.
Some users argued that opting out is practically impossible. “It’s impossible to prevent your work from being scrapped unless you never publish digitally,”one wrote. “Pirate libraries prove it—if you’ve made something, it’s already in the dataset.”
NSFW frontiers
As legal debates unfolded, users tested the platform’s NSFW limits. Within hours, X feeds filled with AI-generated: Commercials for sex toys, complete with glossy cinematography. Trap anime romances exploring queer relationship tropes. Festival scenes like “Sora Bacchanalia,” where toga-clad revelers danced around fires and poured wine over feasts, bypassing Sora’s anatomical censorship filters—designed, apparently, to yield “Barbie doll” nudity.
The veteran “jailbreaker” Pliny also documented a Sims-like sex scene overlay.
A new era of synthetic everything
Sora 2’s audio engine, cameo system, and opt-out IP policy revealed a broader direction for OpenAI: synthetic media as a platform, not a novelty. But the launch’s viral aftermath underscores how quickly the technology outpaces both legal frameworks and cultural norms.
In 24 hours, Sora turned social media into a mass participatory remix engine—collapsing the boundaries between parody, identity theft, and fandom.
i get the vibe here, but…
we do mostly need the capital for build AI that can do science, and for sure we are focused on AGI with almost all of our research effort.
it is also nice to show people cool new tech/products along the way, make them smile, and hopefully make some… https://t.co/bcCUmXsloP
— Sam Altman (@sama) October 1, 2025
Whether this represents the dawn of a creative renaissance or a copyright free-for-all, one truth is clear: AI video no longer needs reality’s permission.
Apparently, sitting atop a $500 billion company makes you immune to public clowning. “Not sure what to make of this,” Altman conceded after watching the deluge.
Apparently, more money.
Generally Intelligent Newsletter
A weekly AI journey narrated by Gen, a generative AI model.
Source: https://decrypt.co/342593/openais-sora-2-unleashed-internet-chaos-dildo-ads-furry-ceos