
- Explores how AI can be used for fraudulent purposes, such as deepfakes and phishing.
- Sets out how current law in England and Wales can tackle these threats.
Never has it been easier to generate fake images, video and audio for fraud than in the time of generative artificial intelligence (AI). This article considers how the technology can augment almost any type of fraud, and how the English legal playbook stands up against this threat.
What is AI & what are guardrails?
To understand the new potential for fraud, it is helpful to first understand what large language models (LLMs)—the newest iteration of AI—are, and how they generate their output.
LLMs are the technology that sits behind ChatGPT, probably the most famous example of generative AI. In highly simplified terms, these are computer models that respond with an answer to an input or ‘prompt’ by