Ask any law firm leader whether their people are using AI, and the answer is almost certainly yes. Ask them whether they're confident it's delivering value, and the room gets quieter.
That gap is the headline finding of our new report, State of AI Readiness in Legal 2026, based on a survey of 160 legal professionals across the US and UK. It's the clearest evidence yet that AI adoption in law firms has outpaced the governance and knowledge infrastructure needed to use it safely - and the data is more candid than many firms may let on.
87% of respondents say they're using or experimenting with AI in some form. That number alone would suggest legal is further along than most industries. But dig one layer deeper and the picture changes: only 14.4% are very confident AI is delivering real value, and 52.5% are not confident or only slightly confident.
Put plainly, most firms have turned the tools on. Very few are sure they're working.
That uncertainty shows up as risk, not just doubt. 65.6% of respondents say their organization is not ready, or isn't sure it's ready, to scale AI safely. And when asked what worries them most, 58.8% pointed to a client-facing mistake caused by AI - ahead of regulatory breach and data security risk. Meanwhile, 43% aren't confident they even know which AI tools their own lawyers are using day to day. You can't govern what you can't see.
Perhaps the most striking finding is what happens at firms that have invested the most. Knowledge and innovation leaders report a formal AI governance framework at more than double the rate of the broader market - 85% versus 44%, a 41-point gap. On paper, they're well ahead.
But ask those same leaders whether their firm is actually ready to scale AI safely, and the lead nearly disappears: 37% say yes, compared to 33% across the broader market.
"The tools are running. The confidence isn't," says Philippa McIntosh, Lead Marketing Manager at Vable, and the leader of the research behind the report. "Firms that have invested the most in AI are barely more confident about scaling it than firms that haven't started. That tells us the problem was never the technology. It's the knowledge foundations underneath it."
That's the line of the report: AI output is only as reliable as the knowledge infrastructure supporting it. Firms with the budget for governance policies aren't automatically firms with clean, current, well-tagged knowledge for AI to draw on - and that's where the real exposure sits.
The findings speak to two audiences that rarely read the same research: legal professionals managing AI's day-to-day risk, and knowledge and information teams responsible for the systems AI depends on. Both need to see this data, because the report makes the case that these are the same problem viewed from two different desks.
If you're weighing how your own firm compares - on adoption, governance maturity, or where the biggest fears actually sit - the full report has the complete breakdown, including results by firm size and region.
Numbers on a page only get you so far. On Wednesday, August 12, 2026, Vable is hosting Fragmented, Ungoverned, Unconfident - a fast-paced, no-fluff 45-minute online panel digging into what's really happening with AI in legal teams: fragmented knowledge systems, missing governance, and the growing gap between investment and confidence.
If any of the numbers above sounded familiar, this is the conversation to be in.
Media contact: Philippa McIntosh - philippa@vable.com - +44 (0)20 7422 0059