Summize

I get asked which tool to buy all the time. Almost nobody asks who they’re buying it from. I talk a lot about shiny tools & sorting your data out first, because bad in, bad out. Both still stand, but who you're actually buying from gets far less airtime. Every provider will tell you they're a partner & most of them mean it. An F1 car is extraordinary engineering right up until you get in and realise you don't know how to drive it. Legal tech is pretty much the same. The technology matters but it rarely decides whether it works. The questions I'd ask anyone before signing: Who will you actually be dealing with in month three, not the sales team? Is there someone to call who understands legal workflows, not just password resets? Will they help you build a playbook? AI in contracting is only as good as your clause library. What's the plan for the first ninety days? A vendor will describe implementation, a partner will describe what happens after it. Chatting to Alexandria (Lexi) Lutz, GC at contract intelligence system Summize, about exactly this at London’s LegalTechTalk! As a GC she's been on the buying side too. "No in-house team I know started with a playbook. The partner helped us build ours, the vendor sent a template. And after go-live I needed numbers for the board, not anecdotes." Their clients say the same. One team of two said "We're just two people and we need a super strong support partnership", that's the reality for so many in-house teams. It's why who you can call matters more than the demo. The other one is whether it properly sits inside the everyday tools your team already lives in, not 3 million buttons to get there... Another GC, another legal team. They'd bought a contract and matter management tool for efficiency, oversight and, above all, reporting, so they could finally show the business what legal was doing. On paper it was fine. Yet hardly anyone used it, and they couldn't work out why. I asked them to show me their day to day, and it clicked. It didn't connect to anything they worked in, so every email, matter and contract had to be entered by hand. A tool bought to save lawyers time was asking them for data entry. That isn't a training problem. It's a blocker, and no amount of encouragement fixes it. The lesson for anyone buying, ask how it fits the way your team already works before you ask what it can do!

Everyone thinks they need AI. Most teams I meet need structure Ask an in-house team which AI tool they're looking at and they'll tell you. Ask which problem costs the business most and it goes quiet. It goes quiet because teams get marketed features, rarely their own challenges or what solving them would open up. Was talking to Alexandria (Lexi) Lutz GC at Summize at LegalTechTalk. She was also saying how overwhelmed in-house teams are feeling. The market's flooded with AI, the demos all look similar, and teams have almost forgotten why they were looking for tech in the first place, because they had a problem to solve. She's right, and nobody says it out loud. Even working out who to talk to is a job in itself. Then the myths. That AI will fix a process nobody has defined. That the tool with the best demo is the best tool. That buying is the hard part. None of them survive contact with a real legal team. Then the confusion underneath, which is worse. CLM? Contract AI? Drafting? Review? A repository? The categories overlap and nobody explains them. So decisions get made off the demo not a mapped-out assessment. Not because GCs are careless, because the market gives them no neutral way to compare and the demo is all that feels concrete. I've watched capable teams do it. In my experience what most in-house teams are missing is not intelligence. It is structure. Knowing what you have, where it lives, what it commits you to, and being able to answer a question about it without three days of searching. It looks different for every team, which is the point. For some it’s renewals and expiries, the agreement that quietly lapsed. For others its obligations, notice periods, volume commitments, price reviews and auto renewals unopened since signature. For some it's governance, evidencing what was agreed, by whom, under what authority, when someone finally asks. For some it's speed under pressure. A conflict, sanctions, a supply chain failure, and the first question is which contracts are affected. Most can't answer quickly, and slowly is much the same as not at all. For many, it's simply visibility. Sales has a pipeline the CFO can see. What does legal have? If you can't extract it as data, you can't put a graph in a board pack, and legal stays invisible where budget gets decided. Was chatting to a GC who was being pitched contract AI, but the real headache was their main supply agreement. It had expired and nobody noticed until a shipment didn't turn up. No model was fixing that. Knowing what they had and when it ran out would have. And what were they being pitched? A drafting tool, for a team that mostly reviews other people's paper. Sometimes the shiny tool is genuinely the right one. The problem is not the demo, it is not knowing why it is right for you before you sign. Know your problem first and every demo becomes a test, not a pitch. If you're in-house, which of these is costing you most right now?

At this point in 2026, I think we can all agree that most of us have dabbled with AI. What continues to surprise me, though, when I’m sitting around a table with other GCs, is how much of our AI use is still exactly that: dabbling. We use it to draft emails, review contract clauses, or research a high-level legal question to confirm our thinking. But for many legal teams, AI is still being used as an individual tool rather than a capability we are intentionally developing across the function. I’ve been looking for a useful way to help legal teams think about that journey, which is why I was particularly interested in the recent AI Fluency Report from Summize. Rather than focusing on whether Legal is adopting AI, the report focuses on the question I think many of us should be asking now: how do we help our teams become AI fluent? What I particularly like about the report is the framework Summize proposes for thinking about that progression. It moves from individual experimentation, through more structured adoption and integration into everyday workflows, and ultimately toward organizational AI fluency. The report also offers practical ideas for legal leaders looking to help their teams move through those stages. In my opinion, the legal departments that set the bar in the next few years won’t simply be the ones with the highest AI adoption. They’ll be the ones that have built AI fluency across the function. The question for GCs is how intentionally they are leading their teams toward that future. Building AI fluency can’t be left to individual initiative or a handful of early adopters. It needs to become a priority for the legal function as a whole. If you’re thinking about how to move your own team from experimenting with AI to becoming truly AI fluent, I think Summize’s framework is a useful place to start. You can read the full AI Fluency Report here: https://lnkd.in/gDgXGXN5

Contracts contain a huge amount of information about your business. But finding the answers you need can still mean searching through agreements, checking clauses and piecing together information manually. That makes contract intelligence reactive when it could be proactive. Summize’s analytics solution, SIA Analytics, gives teams a faster way to get answers from connected contract data, helping them uncover insights across agreements, clauses and obligations without the usual digging. See what contract analytics can look like when the intelligence is built into the way your teams already work.

Still manually digging through contracts to find answers? Summize’s analytics solution, SIA Analytics, gives teams a faster way to connect dots across agreements, clauses and obligations – turning contract data into intelligence you can actually use. See SIA Analytics in action.

We all need to cultivate a builder mindset in the new age of AI. 🛠️ And no, that doesn’t mean we all need to become developers. To me, a builder mindset is about being curious enough to question how work gets done. What problem are we actually trying to solve? Where does this work fit into a larger workflow? What happens before it? What happens after it? And could there be a better way? Because AI fluency isn’t just about knowing how to prompt. It’s also about becoming a systems thinker. We need to understand the workflows around the work and think about how what we’re building fits into the larger organizational system. Otherwise, we risk using AI to make one task faster without thinking about whether we’re actually making the whole process better. That’s what excites me about this new age of AI. More of us have the ability to experiment, build, and improve how work gets done. But the real opportunity is connecting those individual improvements into systems that benefit all of us. On October 14, I’ll be joining Tom Pratt, Rory Douglas, and Alexandria (Lexi) Lutz to dig into Summize’s AI Fluency Report and what the data tells us about how legal professionals are actually using AI. So please join me! I think it’s going to be a really interesting conversation. 👀 A big thank you to Summize for inviting me to speak and be part of it! 📅 October 14 ⏰ 12 PM ET | 9 AM PT | 5 PM BST Links below to register: https://lnkd.in/epCd7-Cn Explore the Fluency Report: https://lnkd.in/ehvqEcsY