Seeing Is No Longer Believing: Sora 2 Puts a Deepfake Studio in Your Pocket

By the AI Boom Team | Published 12 November 2025 OpenAI’s Sora 2 turns a line of text into a realistic video clip, complete with sound. Six weeks after its launch, those clips are everywhere, and so are the worries.

What Sora 2 actually does

Sora 2 launched at the end of September as an invite-only iPhone app in the US and Canada. You type what you want, and it creates a short video with dialogue, sound effects and surprisingly convincing physics. Its headline feature is called Cameo. It lets you drop yourself, or friends who give permission, into any scene. Every clip carries a visible watermark plus hidden labels saying it was made by AI.

Why critics want Sora 2 pulled

On 11 November, the US consumer group Public Citizen wrote to OpenAI demanding it withdraw the app. The group argues it was rushed out to beat rivals, and that it threatens privacy, people’s control of their own likeness and even democracy, as Euronews reported. Japanese studios and other rights holders have raised copyright concerns too. OpenAI says it is listening to feedback and working with rights holders.

The real risk for business

Fake video of a boss asking for an urgent payment is no longer science fiction. As a result, “I saw it on a video call” is not proof any more. Three cheap defences:

• Agree a code word for any urgent money request.
• Call back on a number you already hold, never one in the message.
• Train your team to pause before acting on any surprising clip.

Our 12-week AI plan covers staff training, and our free tools include a ready-made AI usage policy. Want to understand how AI is changing business, in plain English? Read Get AI Before It Gets You by the AI Boom team, or get in touch at hello@aiboom.uk. © aiboom.uk