Skip to content
Ruling 53 of 61

Creative work · The ruling

Can you use AI to make videos with Sora?

Verdict

No

The verdict

No. OpenAI discontinued the Sora app and website on 26 April 2026 and the Sora API on 24 September 2026. If you made videos there, export them now and use another video tool.

No means: no, or not in a way you should rely on. How we rule

An engraved illustration of a blank sheet of paper with a round terracotta seal, a fountain pen, an envelope and reading glasses on a wooden desk.
Plate A sealed sheet of paper, because every question page ends in a ruling.

Short answer: Sora has gone, so the real questions now are how to rescue anything you made in it and where to make video next. The steps for getting a good clip carry over to any tool.

OpenAI4’s help centre is clear on the dates. The Sora web and app experiences were discontinued on 26 April 2026, and the Sora API, which developers used to build Sora into their own products, was discontinued on 24 September 2026. OpenAI updated that page at the end of October 2026, and its ChatGPT capabilities overview does not list video generation either. So you can no longer make new videos with Sora.

That’s worth knowing before you follow a tutorial. Plenty of guides from 2024 and 2025 still walk you through Sora’s prompt box, remix button and plans. None of that applies now. If a site or app still offers to make “Sora” videos today, check who publishes it before you sign up or pay.

How to do it well

1. Export your old Sora clips first. If you ever made videos in Sora, this is the urgent bit. OpenAI says you can go to sora.chatgpt.com/sunset, click Export, and you’ll get an email when your download is ready. It recommends doing this as soon as possible. If it offers a final export window, it will email you before it starts. After that window passes, OpenAI says it will permanently delete any data associated with your use of Sora. If you paid, refund questions go through OpenAI’s ChatGPT subscription refund route, and any ChatGPT or Sora credits you bought can still be used for Codex.

2. Pick a tool that still makes video. Google’s Gemini app is one option. Its help page says you can generate a video from a text prompt and add up to five images and one video. On a personal account you need a Google AI plan. On a work or school account you need a qualifying Workspace licence. It isn’t available to under-18s, and uploading a video to edit isn’t available in the UK. If you’re starting from photos, making videos from images compares the routes, including a no-AI slideshow edit. If the clips are for a channel, making YouTube videos with AI covers the workflow.

3. Write the prompt like a shot description. Describe one scene: who or what is in it, where it is, what moves and what the camera does. “A slow push-in on a steaming mug on a wooden table, soft morning light” gives a tool far more to work with than “a cosy coffee video”. Keep it to one subject and one movement. Ask for three things at once and you tend to get a muddle.

4. Start from an image where you can. A sharp still you own fixes the look of the clip, so you spend fewer attempts on the style. Upload only images you have the rights to use. Gemini’s help page says so directly.

5. Change one thing at a time. When a result is nearly right, adjust a single detail, such as the lighting, the pace or the camera move, and keep the rest of the prompt the same. That way you can see what each change did. Save the prompt behind any version you like, because you will forget it.

6. Review before you share. Watch the clip twice: once as a viewer, and once looking at hands, faces, on-screen text and objects that quietly change shape between frames. AI video often looks right at a glance and wrong on a second look.

Rules and risks

Clips you already made in Sora. OpenAI’s terms for users in the UK, EEA and Switzerland say that, as between you and OpenAI and to the extent the law allows, you own the Output, and OpenAI assigns you whatever rights it has in it. The same terms warn that other users may receive similar output. That assignment covers OpenAI’s side only. It does not give you rights in a real person’s likeness or in anyone else’s work that appears in the clip.

Copyright in AI video. The Copyright, Designs and Patents Act 1988 has different authorship rules for different kinds of work. Section 9(2)(ab) says the authors of a film are the producer and the principal director. Section 9(3) says the author of a computer-generated work is the person who made the arrangements necessary to create it, but that rule covers only literary, dramatic, musical and artistic works. The Act does not settle how either rule applies to a particular AI clip, so check the terms of whichever tool you use before you build anything commercial on it. For the wider picture, see whether AI art is illegal, and for how one tool’s terms play out in practice, using ChatGPT images commercially.

Real people. Only put someone in a generated video if they know and have agreed. A cloned voice raises its own problems, covered in copying someone’s voice. Sharing an intimate deepfake can be a crime. Section 66B of the Sexual Offences Act 2003 makes it an offence in England and Wales to intentionally share a photograph or film that shows, or appears to show, another person in an intimate state when that person does not consent and the sharer does not reasonably believe they consent. That offence has a reasonable excuse defence and some statutory exemptions. Separate offences in the same section cover sharing to cause alarm, distress or humiliation, or for sexual gratification. Threatening to share such an image is also an offence when the person making the threat intends the person shown, or someone who knows them, to fear it will be carried out, or is reckless about whether they will.

Labelling. OpenAI’s provenance page lists Content Credentials (C2PA metadata) as the signal for its video, and says Content Credentials can be lost through everyday actions such as taking a screenshot or converting a file. So don’t count on an old Sora clip still announcing itself as AI. Platforms set their own rules. YouTube requires you to disclose realistic AI content, such as a real person appearing to say or do something they didn’t, altered footage of a real event or place, or a realistic scene that never happened. It also requires disclosure when AI-generated music is the main focus of the video, so an animated music video with an AI track still needs it. You do this with the “AI use” setting in YouTube Studio when you upload. Apart from that music rule, clearly unrealistic content and minor edits don’t need disclosure. Creators who repeatedly fail to disclose can have a label added for them, content removed or be taken out of the YouTube Partner Program.

This is general information, not legal advice.

Sources

8
  1. help.openai.com1OpenAI Help Center: What to know about the Sora discontinuation
  2. help.openai.com2OpenAI Help Center: Provenance signals in OpenAI-generated content
  3. help.openai.com3OpenAI Help Center: ChatGPT capabilities overview
  4. openai.com4OpenAI: Terms of Use (EEA, Switzerland and UK)
  5. support.google.com5Gemini Apps Help: Generate videos with Gemini Apps
  6. support.google.com6YouTube Help: Disclosing use of altered or synthetic content
  7. legislation.gov.uk7Copyright, Designs and Patents Act 1988, section 9
  8. legislation.gov.uk8Sexual Offences Act 2003, section 66B

FAQ

Questions people ask

Is there a Sora AI app?

Not for making videos. OpenAI's help centre says the Sora web and app experiences were discontinued on 26 April 2026, and the Sora API followed on 24 September 2026. If an app or website offers to make Sora videos now, check who publishes it before you sign up or pay.

Why did Sora shut down?

OpenAI's help page on the discontinuation gives the dates, export steps and refund route, but it does not give a reason. Treat any reason you read in articles or social posts as reporting, since OpenAI's help page does not confirm it.

Can I still download the videos I made in Sora?

Possibly, if you act soon. OpenAI says you can export your content at sora.chatgpt.com/sunset by clicking Export, and you'll get an email when it's ready. It recommends exporting as soon as possible. After any final export window, OpenAI says it will permanently delete data associated with your use of Sora.

Can I use Sora videos I already made commercially?

OpenAI's terms for UK users say that, as between you and OpenAI and to the extent the law allows, you own the Output and OpenAI assigns you any rights it has in it. That doesn't clear rights in anything else in the clip, such as a real person's likeness or someone else's work. The terms also say other users may get similar output, and UK law on who owns computer-generated works isn't settled for AI clips.

How can I use AI to create videos now?

Use a tool that still offers video generation. Google's Gemini app, for example, can make a video from a text prompt plus up to five images, but on a personal account it needs a Google AI plan and it isn't open to under-18s. Whichever tool you use, describe one scene, change one thing at a time and watch the result closely before you share it.

What happened to my Sora credits or subscription?

OpenAI's Sora page sends refund questions to its ChatGPT subscription refund article. It also says any ChatGPT or Sora credits you bought can still be used for Codex.

General information for UK readers, not legal, financial or medical advice. How verdicts are decided.