In-house lawyers turn to ChatGPT over legal AI, new report shows

A new report has revealed that AI use among in-house legal teams is dominated by "non-legal" chatbots, such as ChatGPT, rather than legal-specific AI tools.

Specifically, 66% of teams are using general purpose AI tools - far outpacing the adoption of bespoke legal AI tools, which are used by just 7-17% of teams or less.

The survey, conducted by legal talent company Axiom and based on responses from more than 600 in-house lawyers across eight countries, suggests strong enthusiasm for generative AI, but limited adoption of dedicated legal tools in practice.

The numbers

While 89% of in-house teams said they had increased AI usage in the past year, most are reaching for general-purpose tools. Of the teams using those, ChatGPT leads the pack with 42%, followed by Microsoft Copilot (25%), Google’s Gemini (20%) and Anthropic’s Claude (15%).

What’s more, an alarming proportion of AI use appears unstructured. Only 38% of legal departments reported having an AI use policy, while 36% had provided formal training and less than a third had data privacy rules in place.

The UK picture

The report also reveals that UK legal teams appear to be lagging their international peers. 17% of UK legal teams were classified as "mature" users of the technology - signalling "comprehensive access to and strategic use of AI" - the lowest level of all major markets except Hong Kong and Switzerland.

While 76% of all organisations surveyed had plans to increase AI spend in the future, by an average of 26% (in terms of budget), those figures were only 71% and 22% for UK companies. Hong Kong and Australian-based teams lead the way here.

That said, only 11% of UK-based organisations were classed as "immature" in relation to their AI uptake, one of the lowest across the board, with the vast majority (72%) falling into the "developing" bracket, where tools are still being tested or piloted.