A survey of 120 ‘inhouse counsel and legal leaders’ has found that although there are some speed bumps, adoption of legal AI tools is gaining pace. That in turn promises changing client perceptions and improved internal efficiencies.
For example, the survey by ALSP Factor found that ‘25.3% have already spent between $100,000 to $500,000 on domain-specific legal AI tools; and close to half (43.4%) have upgraded existing legal tech licenses to access particular AI features.’
As with an earlier survey by LegalOn, which found approaching 20% of inhouse teams using a legal AI tool for contract review now, such numbers are impressive relative to where we were before. And that change is all down to genAI.
And, those numbers are only going to rise….which means that the inhouse world’s sense of what is ‘normal’ will change too. How GCs perceive value creation by law firms will evolve, as more routine work is clearly now partly open to rapid automation.
Other findings include that 47.5% of inhouse teams that were using AI tools had built – or more likely had built for them – ‘internal AI interfaces / chatbots’. Or put more simply, they’ve now got a secure connection to ChatGPT, or something similar, with some internally branded UI/UX.
We have seen the same pattern with law firms, many of which quickly spun up a secure ‘portal’ that connected to one or more of the well-known LLMs, but provided security guarantees to enable rapid internal use. Then, they also bought in legal AI tools, and upgraded tools they already had to include new AI features.
In short, inhouse teams are now behaving a lot more like law firms when it comes to legal AI. And that matters, because it’s fair to say – at least from what AL has seen over the years – that inhouse teams have been reticent to really engage with legal tech unless they felt they really had to. This feels much more like a ‘pull factor’ scenario.
Why the change? Probably because genAI provides value so quickly. Need a quick review of a sales contract? Prompt and click. There it is. Need a summary of an internal meeting? Prompt and click. There it is. Need help drafting something? Prompt and click, and prompt and click, and perhaps another click. And there it is. No training done on your part. No complex skills to learn. Just go to the interface, prompt and soon you’re at value.
That said, apparently 33.7% of those surveyed are ‘not confident using enterprise AI tools’. But, not confident covers a wide area. Not confident they can use it? Not confident they are making some kind of regulatory error in using the system? Those are not the same things. Also, which tools? CLM? Point solutions? Something else?
It’s not surprising some inhouse lawyers are nervous, given the amount of misinformation floating around about how LLMs work and what impact they can have.
E.g. plenty of folks tend to think of LLMs as a version of Google, i.e. as a reference site, rather than a tool to apply to their own data. And the way that OpenAI marketed ChatGPT contributed to that way of thinking. People also get confused between something like Perplexity – which is a research site – and the main LLM offerings, which can also regurgitate ‘facts’, but that’s not what they’re best at.
Any road, the key message is: inhouse lawyers are adopting legal AI and they’re doing it at scale – if we extrapolate out the survey sample. This is good news.
—
Advertisement: Legal Innovators California conference, San Francisco, June 11 + 12
