When AI Becomes a Repository of Reputation: Criminal Consequences of False Information About Individuals in the UAE and the United States
Introduction
Artificial intelligence has changed the way individuals search for and obtain information. Instead of searching through multiple websites or sources, a person can simply enter an individual’s name into an AI system and receive an answer within seconds. This raises an important legal question: what happens when the information provided to an AI system is false?
The issue becomes more complex where the same false information is repeatedly provided by different users. For example, ten individuals may independently provide an AI system with the false allegation that a particular person committed a criminal offence. If the system stores and processes this information, a future user may receive the same allegation when searching for that individual. The repetition may create an appearance of credibility or corroboration, even though the information remains false. This raises questions not only of reputation and privacy, but also of criminal liability and responsibility for AI-generated information.
The Digital Repetition Problem
The repetition of false information creates a particular problem for AI systems. Unlike traditional misinformation, which may exist as an individual post or statement, AI can process information from multiple sources and present it directly in response to a person’s name.
The legal issue is therefore not simply whether the original allegation was false, but whether its subsequent processing and communication by an AI system constitutes a further publication or dissemination, and who should be responsible for it.
Criminal Consequences in the UAE
The UAE’s cybercrime framework provides several provisions that may be relevant where false information about an identifiable individual is communicated through an AI system. Of particular relevance, Article 43 of Federal Decree-Law No. 34 of 2021 on Countering Rumours and Cybercrimes criminalises using an information network, information technology means or an information system to insult another person or attribute to them an incident that may expose them to punishment or contempt by others. The provision provides for imprisonment and/or a fine not less than AED 250,000 or more than AED 500,000, subject to its requirements.
This provision may be relevant where a person knowingly provides an AI system with a false allegation that another individual committed a criminal offence. The fact that the allegation is communicated through an AI platform does not necessarily remove the original conduct from the scope of existing cybercrime legislation. Article 44 may also become relevant where private or personal information is published with the intention of causing harm, regardless of whether the information is true or false. Furthermore, Article 52, concerning the dissemination or circulation of false news, data, rumours or reports, may apply in circumstances falling within its specific requirements.
A further issue concerns establishing criminal responsibility for an AI-generated allegation. Where a user knowingly provides false information, responsibility may be more readily attributed to that individual. However, the position becomes less clear where the user provides information without knowing that it is false and the AI system subsequently combines or transforms that information into a false allegation. In such circumstances, liability requires consideration of the parties’ respective conduct, knowledge and intention, as well as the causal connection between the original information and the final AI-generated statement.
The United States Approach
The United States presents a different legal framework. Defamation is primarily governed by state law and is subject to constitutional protections under the First Amendment. In New York Times Co. v. Sullivan, the Supreme Court established the actual-malice standard for certain defamation claims involving public officials.
The US approach also raises questions concerning intermediary liability. Section 230 of the Communications Decency Act of 1996 generally protects providers of interactive computer services from being treated as the publisher or speaker of information provided by another content provider. However, generative AI creates a more difficult question where the system does not merely reproduce user-generated content but processes information and generates a new response. Unlike the UAE, defamation in the United States is generally treated as a civil rather than criminal matter, making the consequences of false AI-generated allegations significantly different between the two jurisdictions.
Who Is Responsible for an AI-Generated Allegation?
The central legal problem is determining responsibility where information passes through several stages: a user provides information, an AI system processes it, and another user ultimately receives the resulting statement.
Where the original user knowingly provides a false allegation, existing criminal law may provide a clearer basis for liability. However, where the AI system independently generates the false allegation from information supplied by multiple users, it becomes more difficult to identify the human actor responsible for the final statement.
This creates a potential gap between traditional criminal law and AI-generated content. Existing offences generally focus on human conduct, knowledge and intention, while AI systems can process information from multiple sources and produce statements that were not directly written by any individual.
Conclusion
The use of AI to collect, process and communicate information creates new challenges for existing concepts of reputation, publication and criminal responsibility. The UAE’s cybercrime framework provides provisions that may apply to false electronic allegations, particularly where a person knowingly attributes criminal conduct to another individual. However, the more difficult question is whether these existing laws are sufficient where the final allegation is generated by an AI system rather than directly made by a human.
The comparison with the United States demonstrates that the legal consequences can differ significantly depending on the applicable framework. Ultimately, the key issue is not simply whether AI can generate false information, but who should be responsible when an AI system communicates false information about an identifiable individual.
Copyright: Doctor Mohammed Hassan Al Raeesi Advocates & Legal Consultants retains all intellectual property rights to this content. No third party may use, copy, or modify any part of it without prior permission and without proper attribution to our firm.
This article is for general information purposes only and does not constitute legal advice. The applicable law and procedures may vary depending on the circumstances of each individual case.