How many times have you seen the hype of AI launch with big energy and then quietly fizzle out?
That seems to be a common thread with AI projects in the legal industry (among others).
There’s a new AI tool, a demo or trial, some loose conversations about efficiency. Then the project drops before it ever becomes a daily tool.
That doesn’t mean that AI lacks value. It points to something else.
Despite the fact that budgets are up and AI spending keeps increasing, a recent report on enterprise AI found that about half of all AI projects are stuck in a proof of concept state.
The interest is there, but the movement is not.
The common thread is not resistance to technology, it’s the uncertainty behind it.
Clarity Uncertainty
Most AI projects start with a general concept like “use AI to save time” but then lack defining where that time should be saved, how it will do so (and success be measured), and who owns the process for doing it.
This lack of focus makes the entire thing unclear, and things start to drift.
People test tools, they compare outputs, and trade ideas. But then nothing really gets approved for use because there’s no clarity on what “ready” looks like.
Security Uncertainty
Then the concerns about security and compliance kick in. That throws up roadblocks faster than any lack of project management focus.
And rightfully so. If you don’t hesitate here you might have other issues to deal with.
Law firms handle confidential information, privileged communications, and sensitive internal data.
Before any tool (AI or not) becomes part of the daily workflow, the data handling needs to be understood and management decisions need to be made.
What data can and cannot go into it? Where is it stored? Who reviews the output? How is usage and access monitored? If any of these questions are not established, that throws up a big stop sign to any project.
Skill Uncertainty
It may not be obvious at first because AI tools seem so easy to use, but the skill to use the tool needs consideration. There’s a training/learning curve process, an evaluation of output, reliability, and manager oversight to step in when things drift off track or miss target results. What’s the standard operating procedure around using the AI tool?
How to Plan the Solution
The solution I recommend is to take a narrow approach. Don’t try to change everything or save the world. Just start with one specific use case to implement, then expand from there.
Pick something simple and finite. Something like internal summarization, transcript review, document organization, knowledge retrieval, intake support.
Just one process that is contained, useful, and measurable.
Also make sure to establish your boundaries early. These rules will help you implement because it will be clear what is allowed and not allowed. It will also help you measure your results.
What should the tool do? When do people need to review the results? What types of things are off limits?
Then review.
If your AI project improves workflow, saves time, maintains quality, and fits your risk standards, then congratulations! The hard part is over. You can look at rolling it out officially and consider what your next AI project will be, scaling it up slowly one tool at a time.
Remember that AI projects don’t stall out because the technology isn’t ready or that the technology is too advanced. It stalls because the goals are too vague, guardrails undefined, and nobody is certain of the process.
Starting simple and choosing one practical use case, establish safeguards, keep humans involved, and measure the result against a real business outcome will give your far more success than trying to revamp everything with the promise of AI.