Company legal teams find new ways to use AI tools at work

In short: More company legal teams are using AI tools in practical, sometimes unexpected ways to save time and reduce risk.

What's going on

Company lawyers who work inside businesses, not at law firms, are increasingly experimenting with AI. Instead of buying one ready-made product, some are building small custom systems that fit their day-to-day work.

At Japanese fintech Upsider, the top lawyer Kensuke Tsujimoto is developing a tool that sends workplace conversations from apps like Slack and Notion into Anthropic’s Claude. Claude is a “generative AI” model (software that writes and summarizes text, like an autocomplete that can handle full documents). It flags possible legal issues, like contract problems or rules the company may be missing, and Tsujimoto then checks whether the warning is real.

Other teams are repurposing tools, a bit like using a Swiss Army knife for a job it was not originally designed for. Cox Media Group said it adapted a contract-review tool to help review political ads, which can bring defamation risk if claims are not checked carefully. Nanyang Technological University in Singapore used a contract platform from Lexagle to improve how it handles reports of wrongdoing, and said average case closure time fell to 38.5 days from 55 days.

What to watch

Budgets are not rising as fast as demand, so legal teams may try to get multiple jobs done with one AI tool. Experts warn that just because a tool is approved for one use, it might not be safe for another, and mistakes can create legal risk. For now, many companies will keep humans in the loop and may still rely on outside specialists while they test what works.

Source: Financial Times