Brands Deploy AI-Generated Influencers on Social Media

According to an investigation by The Guardian, brands and marketers are deploying AI-generated influencers and synthetic customer content on major social platforms without clear disclosure. The Guardian reports that some creators producing AI influencer content have been asked to sign non-disclosure agreements, preventing them from speaking about the work. The Guardian cites analysis by the deepfake-detection group Reality Defenders that flagged examples, including imagery connected to the photo app Once.

The consumer group Which? told The Guardian that its testing found 70% of people were unable to correctly identify all the real and fake videos shown to them, a statistic the group used to argue for clearer labelling of AI-generated promotional material. The Guardian also reports there are no specific UK rules that require brands to disclose AI-generated advertising. The investigation notes that the European Union Artificial Intelligence Act will require labelling of AI-generated or manipulated images, audio and video starting in August, but that those requirements will not apply in Great Britain, per The Guardian. Reporting by PYMNTS adds that the U.S. Senate Judiciary Committee voted to advance the NO FAKES Act, legislation that would create intellectual-property protections for a person's voice and likeness.

Industry reporting and vendor guides make clear that what the public calls "AI influencers" covers a range of methods: generative-image models and CGI for visuals, audio synthesis for voice, and scripted or ML-driven persona management for posting and engagement. Sources such as Kolsquare, Metricool, and CreatorDB/Boston 25 describe virtual influencers as programmed digital personas that can be fully produced with synthesis tools and managed at scale. Observed patterns in comparable deployments include tight creative control, repeatable imagery, and the ability to position virtual creators in engineered scenarios that are difficult to reproduce with human talent.

The story highlights two practical frictions for practitioners and platforms. First, content-detection and provenance tooling are still evolving, so brands and platforms face growing public scrutiny when synthetic media is used without disclosure. Second, the regulatory landscape is fragmenting: The EU AI Act introduces mandatory labelling in the EU from August, while the UK currently lacks parallel mandatory disclosure rules, and U.S. legislative efforts such as the NO FAKES Act are at the committee stage, according to PYMNTS. These differences create compliance complexity for cross-border campaigns and for practitioners building detection or compliance features into ad pipelines.

• •Regulatory timeline: whether the UK adopts labelling rules aligned with the EU AI Act, and the legislative progress of the U.S. NO FAKES Act (PYMNTS reporting).
• •Detection and provenance tools: validation work by groups such as Reality Defenders and the adoption of watermarking or metadata standards by platforms and advertisers.
• •Contracts and disclosure practices: whether non-disclosure provisions around AI-created content become more common, and how platforms respond to undisclosed synthetic promotional posts.

Practitioners building content pipelines, moderation tools, or media-audit systems should monitor detection-accuracy baselines from independent groups and the evolving labelling rules in jurisdictions where campaigns run. Industry observers will also watch platform policy changes that affect monetization or ad certification for synthetic media.