AI Haven
NSFW AI

SpaceXSI Argues Grok Deepfakes of UK MP Are 'Obviously Fictitious' โ€” So No Privacy Breach

SpaceXSI (formerly xAI) told London's High Court that Grok-generated sexualised deepfakes of UK MP Jess Asato are 'obviously fictitious' and can't breach privacy.

AI Haven NewsPublished November 21, 20263 min read3 cited sources

Key Facts

  • 1SpaceXSI filed its written defence in London's High Court on October 29, 2026
  • 2UK Labour MP Jess Asato filed the lawsuit in July 2026 over non-consensual deepfake images and videos
  • 3SpaceXSI argues the images are 'obviously fictitious' and 'AI-generated and implausible'
  • 4The company removed a video depicting Asato being chloroformed and prepared for sexual assault
  • 5Asato is seeking damages, removal of images, and a court order for permanent technical measures
  • 6A separate lawsuit alleges Grok generated 7,000+ CSAM images from a single childhood photo

SpaceXSI Files Formal Defence in Landmark Deepfake Lawsuit

SpaceX Super Intelligence โ€” the Elon Musk company formerly known as xAI โ€” has told London's High Court that sexualised deepfake images of a British MP generated through its Grok chatbot are "obviously fictitious" and therefore cannot breach her privacy. The defence, filed on October 29 and reported by Reuters on November 19, marks the first substantive legal test of whether an AI provider can be held liable under UK privacy and data-protection law for user-generated deepfake content.

Labour MP Jess Asato, who represents Lowestoft in Suffolk, filed the lawsuit in July 2026 after manipulated images and videos depicting her in sexualised and degrading situations โ€” including a video showing her being chloroformed and prepared for sexual assault โ€” were created and shared online using Grok. She is seeking damages, removal of the remaining images, and a court order requiring permanent technical measures to prevent Grok from generating such images of her.

The 'Obviously Fictitious' Defence

SpaceXSI's written defence, filed in London's High Court on October 29, argues that the images it left available were so "obviously fictitious," "AI-generated and implausible" that a reasonable viewer would recognise them as fake. The company contends that Asato therefore had no reasonable expectation of privacy in relation to those images, and that they could not breach her rights under UK data-protection law.

The company removed material that violated its terms of service โ€” including the chloroform assault video โ€” but disputes responsibility for other images that remained available. SpaceXSI also denies liability for users' prompts, arguing it "did not determine the content of the relevant users' prompts."

Asato's legal team has argued that Grok's design and training intentionally enabled users to generate non-consensual sexualised deepfakes, making SpaceXSI responsible for processing and misusing her personal data. In comments reported by the BBC, Asato said the company's defence effectively argues she "should reasonably have expected" to be deepfaked.

What This Means for NSFW AI and Deepfake Liability

The case tests a question that the NSFW AI industry has been watching closely: can a platform that permits adult content โ€” Grok allows NSFW output through its "spicy mode" and age-gated settings โ€” escape liability by arguing that obviously fake content cannot harm real people?

If SpaceXSI's defence succeeds, it could set a precedent that AI platforms bear minimal responsibility for user-generated deepfakes of real people, as long as the output is visually identifiable as AI-generated. If it fails, AI companies offering NSFW-capable models may need to implement far stronger identity protections, name-blocking, and likeness filters โ€” similar to the technical measures Asato is requesting.

The case also intersects with broader regulatory trends. The UK made creating or requesting a non-consensual deepfake image of an adult illegal under the Online Safety Act, and the EU AI Act's transparency requirements took effect in August 2026. A ruling against SpaceXSI could accelerate calls for AI-specific liability frameworks across Europe.

Context: A Growing Legal Storm Around Grok

This is not the only deepfake lawsuit facing SpaceXSI. In August 2026, a fourth plaintiff joined a separate lawsuit alleging Grok generated over 7,000 CSAM images from a single childhood photo. The company also faces a Minnesota lawsuit over the state's first-in-nation nudify ban, which a federal judge declined to block in August.

Asato's case has not yet reached a final hearing or judgment. SpaceXSI's defence was filed on October 29, and the company's legal team has not commented publicly beyond the court filings. The High Court has not set a trial date.

Why It Matters

This case will help determine whether NSFW AI platforms can be held liable when users generate non-consensual deepfakes of real people through their systems. If SpaceXSI's 'obviously fictitious' defence succeeds, it could weaken privacy protections for anyone whose likeness is used in AI-generated adult content โ€” directly affecting companion and NSFW AI users who may be depicted without consent.

Sources & Citations

3 cited
  1. 1Musk's SpaceXSI denies liability in UK lawmaker's lawsuit over Grok imagesReuters ยท rmb.reuters.com ยท 2026-11-19T00:00:00.000Z
  2. 2Suffolk MP says Grok was designed to make sexualised imagesBBC News ยท bbc.co.uk ยท 2026-11-20T00:00:00.000Z
  3. 3MP says Grok was designed to make sexualised imagesYahoo News (via PA Media) ยท ca.news.yahoo.com ยท 2026-11-20T00:00:00.000Z