The AI Paradox Coming for Legal, and How GCs Survive It.
“We adopted AI to reduce work. Somehow, we have more!”
I’ve been hearing that line more and more lately in my discussions with legal leaders. Sometimes with a smirk, sometimes with a grimace. Either way, it’s delivered with a healthy mix of irony and dread.
What we’re starting to see as the legal industry dives deeper into AI is the first taste of Jevons Paradox — the idea that increasing efficiency doesn’t always reduce workload. It can do quite the opposite! Efficiency can shift people’s expectations, encouraging them to ask for more, faster, and with higher expectations.
And unless GCs get ahead of it, AI will not close the gap between legal demand and capacity. It’ll blow that gap wide open and drown your team into the wrong kind of work.
The Hidden Ways AI Adds Work
AI is transforming legal. Let’s not pretend otherwise.
But the teams actively instituting AI into their workflows? They’re the ones discovering the little pitfalls no one talks about in the demos. It’s not the technology that’s tricky — it’s what that technology unleashes.
Infinite versions = infinite review loops
When it’s free to generate multiple contract drafts, people will generate them. Why? Because they can. Great. Now you’ve got more drafts to review. (Have fun with that!)
Cheap tasks create dumb tasks
Someone gets the bright idea to run AI over your entire contract archive “just to see what’s in there.” Spoiler: it’s mostly stuff no one needed to know. It’s too easy to get distracted by tangents with no ROI.
You still need a human in the loop
The hallucination risk is real, and no one wants to be the first GC hauled in front of the board because their AI invented a clause out of thin air. Every output still needs a pair of human eyes on it. And those eyes, last I checked, cost money and time.
Everyone thinks legal is now instant
The second someone in the business hears “AI can do it in seconds,” your inbox becomes a vending machine for redlines. Requests go up and response expectations go off the charts. It’s Uber Eats, but for legal advice.
Operational lift gets heavy, fast
This stuff doesn’t install by flipping a switch. There’s integration, security, and training. And just when you’ve got it sorted, the AI vendor launches a new feature, bringing you back to square one.
Bottom line: AI can create great efficiencies and speed up your workflow. But it can also invite the revenge of unintended consequences. Consequences legal doesn’t need when CFOs are already scrutinizing your every dollar.
The Way Out: Smart, Disciplined Implementation
At the end of the day, AI isn’t the problem. Undisciplined AI is.
Smart GCs recognize the difference between reacting and responding is time. Slow down the hype cycle and do what legal has always been good at: applying judgment.
- Start (ruthlessly) small
One use case. Maybe two, tightly scoped and with clear goals. Once you’ve found actual value, then scale. - Set boundaries early-on
Sometimes, what’s best for the company is actually to have legal be the “department of no.” Legal isn’t customer support. You’re not here to respond to every spur-of-the-moment prompt someone throws at you. - Coordinate with your firms — or risk doubling the work
AI makes it easy to crank out analysis, summaries, contract markup, etc.. But if your firm’s using AI to draft the memo and your in-house team is doing the same… congratulations, you just doubled the inputs on the same problem. - Measure value like a CFO
“Efficiency” is not a measurable result. “Hours saved,” “spend avoided” is a result. Get specific with your metrics. Otherwise, your AI strategy is a science experiment, not a business case. - Focus less on what legal does. Focus more on how legal works.
The GCs who are unlocking real value aren’t just automating busywork — they’re automating decisions. An AI legal procurement assistant can help you determine the best firms to engage, when, and at what price.
AI as a Business Force
At our recent Exchange London event, Thomas Laubert (GC at Bayer) nailed it:
“AI has such an impact we have not seen before. Plus there are very few companies who are not under financial stress.”
AI is arriving right as companies are more focused than ever on the bottom line. There’s a temptation to believe AI is a savior, but the devil is definitely in the details.
In a pressure cooker like today’s legal department, these details can backfire if you don’t look out for them, or worse, if everyone else assumes you do.
We’re not in the sandbox anymore. This isn’t about experimenting with ChatGPT over coffee. This is about AI restructuring how legal gets done.
That’s why I’m excited to be moderating our upcoming webinar: Leading Through the Age of Agentic AI webinar on December 9. Thomas Laubert and Lewis Bretts (Partner at PwC) will discuss what’s working (and what isn’t) when AI gets real inside legal departments.
Join us. Because the GCs who thrive in this new landscape will be the ones who gain efficiencies with AI – without erasing those efficiencies with all the extra work.
Cheers,
-Jim
This article originally appeared in the Value Standard Newsletter, which delivers fresh benchmarking, AFA expertise, and insider insights from Fortune 500 teams to your inbox every other week. Subscribe here.
