Mintz On Air: Practical Policies — Trust, Tone, and Trade-Offs: Negotiating Contracts in the AI Age

What role should AI play in commercial contract negotiations, where is human judgment still essential, and how can you continue to build trust?

For in-house counsel, contract managers, and legal operations professionals navigating those questions, the latest episode of Mintz On Air: Practical Policies offers guidance on using AI effectively while recognizing its limitations.

Host Jen Rubin is joined by Tali Tuchin to discuss:

• How AI is impacting contract negotiations
• Where AI might miss important nuance, the assessment of legal and business risk, and critical marketplace experience and assessment
• Why human judgment and skill remain critical to building trust in negotiations
• How best to leverage AI in commercial contract negotiations

Mintz On Air: Practical Policies — Trust, Tone, and Trade-Offs: Negotiating Contracts in the AI Age Transcript

Jen Rubin (JR): Welcome to Mintz On Air: Practical Policies. Today’s topic: Trust, Tone, and Trade-Offs: Negotiating Contracts in the AI Age. I’m Jen Rubin, a Member in the Mintz Employment Group with the San Diego–based Bicoastal Employment Practice, representing management executives and corporate boards. I am always looking to provide creative solutions to workplace problems. Thank you for joining Mintz On Air. If you have not tuned in to our previous episodes and would like to access our content, please visit us at the Insights Center at Mintz.com, or you can find us on Spotify. 

Today I’m thrilled to be joined by my San Diego–based partner, Tali Tuchin. Tali has over 20 years of experience managing a wide range of complex transactions for businesses in the biotechnology, pharmaceutical, diagnostic, medical device, health care, and energy sectors. Tali provides counsel on collaborations, strategic alliances, patent licenses and assignments, research and development agreements, and joint development agreements. Her focus is on keeping critical deals on track, working efficiently, and maximizing business value. Tali draws on her prior experience as an in-house corporate counsel for a large, publicly traded biotech company to serve as a trusted business advisor. Companies rely on Tali’s comprehensive transactional experience and insightful advice to manage investor expectations, control legal costs, and address the needs of diverse internal stakeholders and navigate negotiation challenges.

How AI Is Changing Commercial Contract Negotiations

JR: Tali, I’m betting AI has impacted your approach to negotiating commercial contracts in more than a few ways. Am I right about that?

Tali Tuchin (TT): Oh, yes. Absolutely. 

JR: Let’s set the stage for today’s discussion. Let’s imagine a world where a complex commercial agreement — let’s call it a 50-page agreement, with all sorts of complex licensing, indemnity rep, and warranty issues — is presented to you. I recognize, Tali, this is your bread and butter, but why can’t you just upload that agreement into your AI system and tell it to flag the important issues and make suggestions on how you should fix them? I realize this is a simplistic scenario, but I need to set the stage.

TT: I love that hypothetical. It reminds me of saying to an AI architect, “Please build me a house.” Yes, you might get something, but obviously we have to back up. We have to start with, what does the client want? What are the client’s objectives and goals? What items are important to the client in this transaction? What are they trying to get out of this deal? Part of that includes, what are the risks of this deal? What are the potential situations down the road that could cause things to go sideways? How do we capture those risks?  We also have to know things like the party dynamics. This is a negotiation. This is between two entities. What is the relationship between them? What is the history? 

Those are things that obviously aren’t able to be fed into your AI. So much takes judgment, not only in negotiating, but in drafting. You have to use your judgment as a lawyer. Part of that judgment is based on your years of experience and exposure, and having seen all flavors of negotiations, having seen all sorts of signed contracts be litigated over or disputed later. Your judgment as a lawyer is a key issue in putting together that contract or building that house. 

JR: I love that analogy. Get the client the house it wants. But how do you know what house it wants? Well, you have to ask questions. You have to build the relationship. There’s value in learning how to communicate with your client and understanding what your client is trying to achieve. But let’s give AI its due. AI definitely does some things really well, right? 

TT: Absolutely. It will save you a lot of time. It can take away some of the grunt work that nobody likes to do. For example, putting together a playbook for a straightforward agreement where you would like to have fallback provisions and don’t necessarily want to call your lawyer — let’s say for a nondisclosure agreement. You can help your AI by building precedent or examples in a library, and you can ask the AI to pull from that. It can even produce a workable initial markup of some basic NDAs. 

But as you know from other AI conversations, it does fabricate things. It makes assumptions and it misses things. It is, after all, predictive. So part of what’s tricky is making sure the AI can identify — and confirming that it has identified — any deviations between where, say, a counter-party’s draft agreement departs from the client’s standard positions or departs from an agreed term sheet. The AI tool can rank these changes, and if you inform it what provisions are most material to your client, then it can help rank the materiality, too. The other thing AI can do pretty well is quality control, like checking internal cross-references and spelling and things like that.

The Limits of AI in Contract Drafting and Legal Risk Assessment

JR: So it’s an automation issue, right? AI adds that automation. But there are limits to automation. Sometimes there are things humans just do better that cannot be automated. In terms of judgment calls and nuance, what do you think about that, Tali? 

TT: Part of that goes back to the foundational points you made, Jen, which is, we have to find out what is the specific situation for the client and the other company, the purpose of the deal, the big picture of the deal. The AI might not know, for example, that this is a key deal that needs to be press released before a trade show, or that this deal has to come in under the wire before an SEC filing. That is the kind of nuance AI doesn’t understand. 

Another thing I’ll point out is that part of what we need to do when we’re putting together agreements — whether drafting them from the start or asking the clients questions — is we have to be able to see into the future as best we can. Part of a good contract is making sure you’ve tried to identify all the possible future situations, and then account for them. What if this, what if that? Even if you feed AI as much detail as you can about the deal, it still can’t look around corners as well as a seasoned, experienced lawyer. 

Why Human Judgment and Negotiation Experience Still Matter

JR: All of the things you see in your practice as a commercial lawyer informs your judgment. It makes us look at things and view them through a different lens, because we have that experience. And we are very much not robots, right? We know how to gauge that human intelligence factor. We know how this might fit within the industry or within the market knowledge. How does that impact your relationship with your clients, and how you proceed? AI is a tool that clients are expecting us to use these days. How do you talk to your clients about those things? 

TT: Well, they should be expecting us to use it. It’s as fundamental as having access to the internet or a spell check. Every good lawyer should be using a robust, trusted AI tool as an assist. But as you said, there are nuances. There are the vagaries of human communication and the need for background and experience. 

Going back to the complex agreement you teed off with: Let’s not forget, once you write it, and even if you have the help of AI and you know it’s watertight and everything is accurate, this is going to turn into a negotiation at some point. Negotiations are human, face-to-face interactions. They require quick thinking, quick responses. If you don’t have total authorship of your contract, you might not know it very well and you might not know the why of that contract. If you’ve delegated too much to your AI tool, you’re not going to be in a strong position to negotiate and defend that contract. And you’re certainly not going to be in a good position to respond to the other side as they poke holes or challenge or ask you to defend those clauses. So this human interaction, I don’t think AI has solved that yet. 

JR: What we’re talking about is credibility, right? When you’re negotiating something with opposing counsel, you’re floating things as trades. And AI doesn’t have that memory of those relationships and credibility — and frankly has no reputation to protect. How you behave in the marketplace is a reflection on your client. That to me is one of the most important things lawyers can take away. That is the human factor that is so important.

Also, of course, the AI can’t read the room, right? How many times have you encountered the same lawyer on the other side where you’re able to utilize and leverage that credibility you built with that person? Talk a little bit about that. 

TT: Oh, absolutely. Ideally you develop a relationship of trust and mutual respect. And that is what your client wants out of a contract negotiation. A contract isn’t a litigation. Both parties want to conclude the deal and get going with their business. Clients want us to do this as efficiently and with the least expense as possible. To be efficient, you need to create trust. You need to create a collaborative environment. You need the other side to know that you’re listening to them and you’re hearing them. And that, of course, includes listening to the other company, not just the other company’s lawyer. What is the other company concerned about? What are they afraid of? The way you arrive at a contract that meets everybody’s needs is by accommodating, as best you can, each party’s goals and risks. A lot of that also comes back to human interaction, nuance, experience, understanding from prior deals what could be a creative solution. What can overcome a negotiation hurdle or obstacle? I haven’t seen AI do that either. 

The other thing I want to mention is this idea that AI tools can help you understand what’s quote-unquote market or commercially reasonable. Going back to the house example, what is my house worth? What’s market? Every deal, every transaction is different. As attorneys — and certainly as attorneys at a large firm like Mintz, where we have access to a broad variety of resources on the market and we do a lot of research on what is market and we have access to our peers’ and colleagues’ experiences on similar deals — we are able to advise our clients on what is quote-unquote market or more commercially reasonable or standard than an AI tool can. We certainly use our AI tools to do that research, but I wouldn’t trust the AI tool to tell me that myself.

JR: Right. You can’t replace all of the things we see and all the conversations we have, the information we share — because we collaborate here at Mintz, we’re sharing across different practice areas and relating to different industries and different clients. And it does build that aspect of our knowledge database that doesn’t get put into a computer system. These are the things we build amongst each other. 

Building Trust and Collaborative Relationships in Contract Negotiations

JR: I want to go back to the notion of trust. Anybody who’s been following my podcast knows this is something I’m very interested in. In commercial relationships — like employment relationships and any other relationship in our lives — the foundation is trust. Trust informs all the relationships we have and how effective those relationships are. Personal, professional —it all comes back to the same notion, which is credibility. And there’s only so far AI can go in terms of that. 

Let me turn to some of these complex layers. How does AI take optimizing, for example, individual clauses and then give you advice on what to do about those independent clauses? And how do you make use of disparate suggestions? Because that’s happened to me when I’m using AI to give me advice — I look at all the different options, like whether I want the wood floors or the concrete floors. I get lists of recommendations and sometimes I look at them and say, “This doesn’t even make sense.” How do you make sense of all the disparate options that AI gives you in a commercial negotiation? 

TT: You could only really make sense of these things if you have experience with them. So if the AI suggests a certain clause or a certain way of writing a clause and you read it and you don’t understand why it’s written that way, or what the goal is, or what the effect of that language is, you’re not in a position to evaluate that clause at all. In fact, you might be over your skis. That might be a sign of not having the right experience or expertise yet on that topic, and you need to seek out guidance on that, whether it’s from a practical resource or a colleague. But my point is that in order to use AI effectively in any situation, and certainly with contract drafting, you need to know enough about your subject matter to be able to evaluate the practicality and effectiveness and long-term effects of the clause it’s suggesting.

JR: The good thing about the AI tools we use here at Mintz is they do explain the rationale behind what they came up with. So I think the best way to use these tools is to have, as much as you can, a conversation with them, ask lots of questions, probe. And if you don’t understand it, don’t assume it’s right. If you don’t understand it, maybe it’s wrong. And that’s where I call somebody like you or one of my other partners for advice. There’s something comforting in knowing there’s a human I could go to when I know I trust that human, I trust that human’s judgment. 

I pair that with some of the amazing things AI can offer. I’ve been practicing for 37 years, and I would never claim to know everything. But I’ve seen a lot in that time frame, and I still see things that AI suggests to me that I had not thought about. And frankly, between us, I think it makes the practice of law really fun. But that’s a podcast for another day. 

Best Practices for Using AI in Commercial Agreements and Contract Review

JR: I want to bring it back to using AI in a commercial negotiation. Tali, can you give our listeners some bullet points for the best ways to utilize AI in this context?

TT: For counsel who are relying on AI, and not going to outside counsel or outside experts, remember that it is a tool and it is not the final. It can make recommendations it assumes might be applicable to your situation, but those recommendations can be made in a vacuum. Always challenge every provision the AI writes for you.

A client sent me an agreement and he said, “It’s very simple and straightforward, I asked Claude to prepare it. Here is the first draft. Can you work from this?” Number one, it took me longer to fix the Claude agreement than if I had written it myself. But that’s okay. I respected the client’s wishes to start with that. But number two, I went through and made the revisions and tidied it up. I sent it back to the client and he said, “This looks great, but what about this clause, where it says I’m responsible for costs. Why did you put that in there?” And I said, “No, no, that was in there already.” My advice is, if you do send a draft like this to outside counsel, please read it first. Make sure it reflects what you want to see — to the extent you are able to understand and read contract language — before you send it over. I guess it’s not safe for me to assume that the client has read and accepted all of it, but it should be. 

JR: Yeah, I totally get it. Item number one, read any AI-created drafts you’re sending to outside counsel. That’s also super efficient and helps control costs. It goes back to the point you made earlier, which I completely agree with: If you have not written something and you don’t understand it, how do you know what questions to ask about it? You can have all the judgment in the world, but — going back to the house analogy — if you don’t know whether the client’s looking for an ADU or a 40,000 square-foot palatial residence, how do you know what advice to give?  Businesses pay for clarity of thought and industry knowledge. They’re not looking for volume markup. 

Also important is making sure the contract captures the negotiation outcomes. This goes back to understanding your client’s needs. Maybe there are certain periods of time that are important to the client. Maybe there’s something historical going on that, because of your long relationship, you can make sure you put that into context. That develops the superior market intelligence that is specific to what the client needs. The other thing I want to mention is collaboration. A contract negotiation is looking for outcomes that benefit both parties. How do you use AI to help you be more collaborative?

TT: In my experience, AI tools are somewhat more aggressive in terms of what they’re proposing — aggressive, but generic. Positions that are easy to generate. The differentiator becomes the collaborative problem-solving that creates a value. Thinking hypothetically. Being creative. Looking back on your roster of deals and thinking, “This worked in that other deal.” AI is a tool, but it’s just one tool — it shouldn’t substitute for experience and judgment and the human touch. It can’t replace understanding your client and establishing trust with the other side. 

Wrap-up

JR: That’s a great wrap-up here. The real value is in those lawyers who can combine technological fluency with human judgment. We’re taking concepts of trust into consideration, and industry knowledge and experience, and the collaboration that comes with working in a full-service law firm where we bring a lot of different perspectives from a lot of different industries. That’s an inherent value proposition that AI just does not have over us. So there you go. Thank you for joining me today, Tali. Really appreciate it. 

TT: It’s my pleasure, Jen. I’m glad we’re in violent agreement over this. 

JR: For more information about our practice groups here at Mintz and for our other thought leadership, visit the Insights Center at Mintz.com. And you can find our Mintz On Air: Practical Policies podcast on Spotify. I’m Jen Rubin. Thanks again.

Subscribe To Viewpoints