Artificial intelligence is increasingly being used in everyday life, but it is also being misused in ways that can lead to serious criminal allegations. From deepfake imagery to voice cloning fraud and large scale online harassment, the law in England and Wales is evolving to reflect these risks.
There is no single piece of legislation dedicated solely to artificial intelligence. Instead, the courts and prosecutors are applying a combination of newly introduced offences and existing criminal law to address this type of conduct. The result is a developing framework where the focus remains firmly on harm, intent and behaviour, rather than the technology itself.
This article explains the main types of artificial intelligence related offending, the legislation currently being used to prosecute these cases, and how Stephensons can support those facing investigation or criminal proceedings.
Deepfakes and intimate image offences
One of the most significant developments in recent years relates to deepfake technology, particularly where artificial intelligence is used to create sexual or intimate images of real individuals without their consent.
The Online Safety Act 2023, which came into force in January 2024, introduced offences criminalising the sharing or threatening to share intimate images without consent. Importantly, this applies not only to genuine images but also to manipulated or artificial intelligence generated content.
The law has moved further still. By 2025 to 2026, reforms have extended criminal liability to the creation of sexually explicit deepfakes, as well as the act of requesting them. This represents a significant shift, recognising that harm can arise even where content is not distributed.
In practical terms, this means that individuals may now face prosecution for:
- Creating artificial intelligence generated intimate images of another person without consent
- Requesting such images from a third party
- Sharing or threatening to share such material
These offences can carry custodial sentences, reflecting the seriousness with which the courts now treat this type of conduct.
Fraud and artificial intelligence impersonation
Artificial intelligence is also being used to facilitate increasingly sophisticated fraud and impersonation offences. Voice cloning, synthetic video and automated messaging tools can be used to impersonate individuals or organisations with a high degree of realism.
There is no bespoke “artificial intelligence fraud” offence. Instead, these cases are commonly prosecuted under the Fraud Act 2006, particularly fraud by false representation.
Examples include:
- Impersonating a company director using cloned audio to authorise payments
- Creating fake videos to induce financial transfers
- Using artificial intelligence generated messages to deceive victims into disclosing information
The key legal issue is not the technology itself, but whether there was dishonesty and an intention to make a gain or cause a loss. Where those elements are present, existing fraud legislation provides a robust basis for prosecution.
Harassment and malicious communications
Artificial intelligence has made it easier to generate and distribute harmful content at scale. This includes repeated abusive messages, manipulated images, and fabricated allegations designed to intimidate or distress individuals.
Such behaviour is typically prosecuted under:
- The Protection from Harassment Act 1997
- The Malicious Communications Act 1988
These laws criminalise conduct where there is a course of behaviour amounting to harassment, or where communications are sent with intent to cause distress or anxiety.
Where deepfake content is used as part of that conduct, it can strengthen the case by demonstrating intent, impact, and persistence.
The courts will look closely at factors such as:
- The frequency and nature of the communications
- Whether the content is threatening, indecent or grossly offensive
- The purpose behind sending or creating the material
Again, the presence of artificial intelligence does not change the legal test. It is simply another method by which the offence may be carried out.
False communications and harmful misinformation
Another important development is the introduction of offences targeting harmful online communications. The Online Safety Act 2023 includes provisions relating to “false communications”, where a person sends information they know to be untrue with the intention of causing harm.
This is particularly relevant to artificial intelligence generated content, which can create highly convincing but entirely false narratives.
Examples may include:
- Deepfake videos falsely depicting criminal or inappropriate behaviour
- Fabricated audio recordings implying misconduct
- Artificial intelligence generated content designed to damage reputation or wellbeing
In these cases, prosecutors will focus on whether the individual:
- Knew the information was false
- Intended to cause harm
- Lacked a reasonable excuse for sending it
Where those elements are established, criminal liability may follow regardless of how the material was created.
Computer misuse and data related offending
Some artificial intelligence related allegations arise from the way in which content is created or obtained. This includes situations where systems are accessed without permission, or personal data is gathered and used unlawfully.
The Computer Misuse Act 1990 continues to play a central role in these cases. It covers offences such as:
- Unauthorised access to computer systems
- Access with intent to commit further offences
- Unauthorised acts impairing the operation of systems
Where artificial intelligence tools are used in conjunction with unauthorised access or data misuse, these offences may be charged alongside fraud or communications offences.
How Stephensons can help
Allegations involving artificial intelligence are often complex. They can involve large volumes of digital material, technical evidence, and questions around authorship, attribution and intent. Early legal advice is crucial.
At Stephensons, our criminal defence team can assist with:
- Police station representation, ensuring your position is protected from the outset and that you receive clear advice before interview
- Challenging attribution, particularly in cases involving anonymous accounts, shared devices or disputed authorship
- Scrutinising digital evidence, including how material was obtained, analysed and presented
- Advising on intent and knowledge, which are central to offences such as fraud, harassment and false communications
- Preparing a robust defence strategy, tailored to the specific facts, evidence and legal issues in your case
We understand that being accused of an offence involving artificial intelligence can have significant personal and professional consequences. Our focus is on providing clear, strategic advice and strong representation at every stage of the process.
A developing area of law
The legal framework surrounding artificial intelligence is evolving rapidly. While new offences have been introduced, particularly in relation to intimate image abuse, much of the current approach still relies on established criminal law adapted to new forms of behaviour.
What is clear is that the use of artificial intelligence will not prevent prosecution where the underlying conduct meets the legal definition of a criminal offence. The focus remains on what was done, why it was done, and the harm caused.
If you are under investigation or facing allegations of this nature, seeking advice at an early stage can make a significant difference to the outcome. Contact Stephensons today on 0161 696 6188 to speak with our specialist solicitors.
Comments