Deepfake Defamation: How UK Law Tackles AI-Generated Fake Content

Deepfakes — AI-generated images, audio, and video that convincingly depict real people doing or saying things they never did — represent one of the most serious emerging threats to reputation. UK law is adapting to address this challenge, but significant gaps remain.

What Are Deepfakes?

Deepfakes use artificial intelligence, typically deep learning neural networks, to generate or manipulate media. The technology can create entirely fabricated videos of real people speaking words they never said, place a person's face onto someone else's body, or generate realistic but entirely fictional photographs. The quality of deepfakes has improved dramatically — many are now indistinguishable from authentic content without forensic analysis.

While deepfake technology has legitimate uses in entertainment and research, its potential for harm is enormous. A deepfake video of a business leader making racist comments, a professional engaging in criminal activity, or an individual in intimate situations they never participated in can cause devastating and lasting reputational damage.

How Defamation Law Applies to Deepfakes

Existing UK defamation law applies to deepfakes in the same way it applies to any other false and damaging publication. The key elements remain:

• Publication: The deepfake must have been published to at least one person other than the claimant
• Defamatory meaning: The content must lower the claimant in the estimation of right-thinking members of society
• Identification: The claimant must be identifiable in the deepfake
• Serious harm: Under the Defamation Act 2013, the publication must cause or be likely to cause serious harm to reputation

In practice, deepfakes often satisfy these requirements easily. A realistic video of someone committing a crime or engaging in misconduct is inherently defamatory, the person is clearly identifiable, and the potential for serious harm is obvious.

Challenges Specific to Deepfakes

Attribution and Anonymity

Deepfakes are often published anonymously on platforms that do not require identity verification. Identifying the creator may require Norwich Pharmacal orders against hosting platforms, forensic analysis of the content's metadata, and cooperation from technology companies — some of which may be based outside the UK.

Proving Falsity

While it might seem obvious that a deepfake is false, establishing this in legal proceedings can require expert evidence. Forensic analysts who specialise in detecting manipulated media may need to examine the content and provide testimony about the techniques used to create it. This adds cost and complexity to the claim.

Rapid Viral Spread

Deepfakes that are sufficiently convincing can spread virally before the subject even becomes aware of their existence. By the time legal action is initiated, the content may have been viewed millions of times and copied across multiple platforms. Content removal becomes a game of whack-a-mole.

The "Liar's Dividend"

Paradoxically, the existence of deepfake technology creates a "liar's dividend" — the ability for people caught in genuine recordings to claim that the content is a deepfake. This further complicates the evidentiary landscape in defamation cases.

Criminal Law Provisions

Beyond defamation, several criminal law provisions may apply to deepfakes:

• Online Safety Act 2023: Creates a new offence of sharing intimate deepfake images without consent. Platforms have a duty to remove such content
• [Malicious Communications Act 1988](/blog/malicious-communications-act): Sending deepfakes with intent to cause distress may constitute a criminal offence
• Computer Misuse Act 1990: If creating the deepfake involved unauthorised access to computer systems (e.g., stealing images from private accounts)
• Fraud Act 2006: Using deepfakes to deceive for financial gain
• [Protection from Harassment Act 1997](/blog/harassment-uk-law-guide): Repeated creation or sharing of deepfakes targeting an individual may constitute harassment

The Role of AI and Platform Regulation

The Online Safety Act imposes duties on platforms to address harmful content, including deepfakes. Platforms designated as "Category 1" services must have systems in place to address content that is harmful to adults, and all platforms must address illegal content promptly. Ofcom, as the regulator, has the power to impose significant fines for non-compliance.

Several major platforms have also implemented their own deepfake policies, including labelling requirements for AI-generated content and removal of deepfakes designed to mislead. However, enforcement remains inconsistent.

Practical Steps If You Are a Victim

1. Preserve evidence: Screenshot and archive the deepfake content immediately, including URLs, timestamps, and any available information about the uploader
2. Report to platforms: Use platform reporting mechanisms to request removal. Most platforms now have specific processes for AI-generated content
3. Seek legal advice: A specialist solicitor can assess whether defamation, harassment, privacy, or other legal claims are available
4. Consider criminal reporting: If the deepfake involves intimate imagery or was created with malicious intent, reporting to the police may be appropriate
5. Pursue removal: Court orders, including interim injunctions, can compel removal of deepfake content. In urgent cases, injunctions can be obtained within hours
6. Request search engine delisting: Even if source removal proves difficult, search engines may delist deepfake content from results

The Evolving Legal Landscape

The law in this area is developing rapidly. The EU AI Act, which may influence UK regulation, imposes transparency requirements on AI systems that generate or manipulate content. The UK government has signalled its intention to strengthen protections against deepfakes, particularly intimate deepfakes.

As detection technology improves alongside generation technology, courts will become more adept at handling deepfake evidence. In the meantime, the existing legal framework — defamation, harassment, privacy, and data protection — provides meaningful remedies for victims, even if the law has not yet fully caught up with the technology.

Related reading: AI and ChatGPT defamation | Defamation in the digital age | How to remove defamatory content