The use of generative artificial intelligence in legal services this week came under attack – from the founder of a pioneering AI-based law firm.

Raj Panasar, a former partner at international firm Hogan Lovells, described ‘probabilistic’ AI, which constructs statistically likely answers from patterns in data, as ‘fundamentally unsuitable for regulated legal work, where an outcome is either right or wrong’.

Panasar’s startup, LawFairy, was this month authorised as a law firm by the Solicitors Regulation Authority. The firm says its technology is ‘deterministic’ – structured on verifiable rules. This is an apparent return to an older generation of AI, the so-called ‘expert system’. The firm claimed that although it uses generative AI to create a ‘clear interface’, its legal outcomes are produced entirely through pre-validated rules embedded in structured logic. Similar software is deployed in commercial flight simulators, where a given input must always produce the same output and every step of the decision must be auditable. 

LawFairy announced yesterday that its initial focus will be on immigration law. ‘Immigration decisions aren’t about sounding right – they’re about being right, and being able to prove it,’ Panasar said. ‘When decisions affect someone’s right to work, remain, or settle, confidence without accountability isn’t good enough. We’ve built LawFairy so every outcome can show its working.’

The announcement of LawFairy’s approval by the SRA comes as the MoJ-funded LawtechUK initiative said that investment in startups grew by 35% to nearly £200m in 2025.