AI, Enterprises, and the Law

Tom Stoneman:
Hi, I’m Tom Stoneman, and this is The Intelligent Enterprise where every two weeks we take a break from the chaos of enterprise life and get inside a big idea by getting outside of it. Each episode, we meet an industry expert who helps us cut through the noise from all the updates and rollouts while exploring one of their favorite breaktime activities, whatever they do to give themselves a break from day-to-day demands of work and the distance they need to come back to a problem with a fresh perspective. This week, I’m joined by Vintee Mishra, a lawyer specializing in tech law and intellectual property.

Vintee Mishra:
I started my career with a boutique IP law firm, which sharpened my focus towards IP and tech law. And then my career took me to tech [inaudible 00:00:59] and roles that filled the gap between law and what businesses were actually doing.

Tom Stoneman:
Vintee’s currently part of the commercial contracting organization at Navy Federal Credit Union, supporting enterprise scale technology and commercial agreements. Her prior experience includes roles supporting Tata Consultancy Services, Cisco, First Technology Credit Union, and Moody’s Analytics. Vintee joins us today to discuss the exciting ways AI is being implemented in the legal field, but she also discusses the real risks that come along with that mass adoption.

Vintee Mishra:
There’s so much excitement, but there’s a thin line between excitement and overreliance.

Tom Stoneman:
We’ll get into the way AI is boosting efficiency, the ways it’s complicating intellectual property law, and the guardrails that organizations should have to prevent misuse.

Vintee Mishra:
Enterprises will have to get more selective, so the procurement tightens up and legal and compliance get earlier input. They need to be involved from the beginning.

Tom Stoneman:
Let’s get inside the future of enterprises by stepping outside of them. You’re an active advisor on contract lifecycle management platforms with particular interest in how generative AI is transforming contract drafting, review, risk assessment, and legal operations.

Vintee Mishra:
Yes.

Tom Stoneman:
There’s over 500 GenAI powered tools across contracts and litigation and various topics, and there’s more than three tools launching every day. And first pass contract review that took hours before, now it only takes minutes. So my question to you is, how do you feel these tools affect your daily work life cycles?

Vintee Mishra:
Great question. So I am an advisor on the contracts issues to multiple teams across several functions within the organization. And part of the job is definitely to review any AI provisions in the contracts to understand what kind of legal protections we require around it, what is being input, and what is the output, the quality of the output. Do we have enough legal protections like indemnity and liability allocations around it? Until mid of last year, I had heard about over 500 GenAI powered tools across contracts, litigation, research, legal ops compliance, et cetera, like you mentioned. And yes, more than three new tools were launching every single day. I’m sure they have multiplied by now to way more. What I’m hearing from big law firms, from my peers in different industries and organizations, from various subject matter experts, knowledge-based conferences and contract experts, I am hearing that more than 56% of in-house attorneys now regularly use AI in contract review, and this is mainstream.
I’m also hearing that people are advocating use of the AI tools like Harvey or Spellbook in researchers, drafting, issue spotting. I’m hearing most of the major law firms are reporting 40 to 50% faster first draft. So like you mentioned, first past contract reviews used to take hours and now they just take minutes. They are being built-in in the docs, which drives adoption at a much faster rate. I’m hearing that lawyers are saving at least 30 to 50% of time on their contract work. So overall, with all these tools available and all the massive usage we are seeing, there is definitely effect on the daily work life cycle. There has been a massive adoption of Harvey or Spellbook or other tools like Legora, CoCounsel, Clarity, Kira, and so on and so forth. It’s just out there. It’s mainstream now.

Tom Stoneman:
Most of my experience working with legal firms, you’re billed hourly. Is there concern that because AI speeds up the cycle so much that you would have to then take on so many new additional clients or that your overall revenue is going to drop?

Vintee Mishra:
There’s definitely huge concern with the heavy adoption of legal AI tools. About what will happen to the future of legal industry, there’s definitely going to be a massive impact on several legal roles and positions, but at the same time, there will be new opportunities and roles that will get generated and will eventually generate revenues. So as of now, we have to consider AI as a very capable and fast junior associate whose output must be evaluated and reviewed by an expert or a senior attorney. There will be new roles like AI ethics professional, compliance professionals, legal engineers, prompt engineers. So there will always be roles coming in. They’ll replace the typical traditional roles, but revenue will keep coming in because of the faster turnaround time. But what we have to be mindful of the fact is the quality control.

Tom Stoneman:
There’s just so many rabbit holes we could go down. But let’s stick with the enterprise side here. I think this is a question that has come up in almost every one of these podcasts we’ve done talking about enterprise and AI. Do you think there’s an overreliance on AI that there’s trusting without any verification, not enough human interaction?

Vintee Mishra:
There’s definitely an overreliance and it is not because people think that it is the best thing to go forward with. It is mostly because there is a blind race towards it without really understanding the problems they’re trying to resolve. People are jumping to find solutions to remain in the competition without understanding what their problem statement is. There’s so much excitement, but there’s a thin line between excitement and overreliance. A lot of vendors in this space, they claim that they’re highly accurate, but most of those claims are marketing based, not measurement based. And most of the benchmarks or measurement criteria that they have in place is feature based, not entirety of the tool based. And if you’re choosing or selecting an AI tool to, let’s say, do a contract review for yourself, you have to ask questions such as, what are you measuring your accuracy against? Is it the entire end to end contract review, redlining, editing, negotiating, renewals, terminations, et cetera, or is it, let’s say, just issue spotting?
And then there have been so many litigations as well where lawyers have been fined for misleading the courts because they cited inaccurate case laws, unfortunately without verifying or doing the quality check. So globally speaking, AI has hallucinated critical clauses during due diligence. And therefore, like I mentioned before, experienced attorney will have to still validate output, understand context and catch what the model has missed.

Tom Stoneman:
Let’s bounce over to policy, everybody’s favorite topic, at least on the enterprise side. So I know from working in enterprises most of my life that enterprise policies are… It feels like they’re there to make life difficult for us that are trying to work, but they’re not, right? I mean, they’re really there to act as guardrails and to keep everything running smoothly. Do you believe enterprise policies have kept up with AI in the workplace?

Vintee Mishra:
Yes and no. Unfortunately, most policies are written reactively after incidents, and therefore it is hard to say what percentage of organizations have actually kept up with AI in the workplace, but organizations that have defined acceptable use boundaries, which have defined confidentiality and data access controls, which are ensuring that they’re attributing the authorship and ownership clearly, which have created training and awareness programs, which have ensured their disciplinary measures for misuse, and which have involved human intervention for quality control to evaluate AI output are definitely keeping up with the AI in the workplace.

Tom Stoneman:
So with AI changing the value and usage of these business assets, do you see the future of AI and IP shifting from uncertainty towards a more collaborative ecosystem defined by human plus machine, or do you see a more AI going it alone approach?

Vintee Mishra:
It has to go hand in hand. There are bottlenecks and roadblocks because of overenthusiasm and overreliance on AI, and therefore there have been IP litigations. As of March 2025, I think D.C. Circuit ruled that AI cannot be a copyright author, human authorship’s still a legal requirement. As far as the copyright, landscape is concerned. As of now, works created entirely by AI with no human contribution, cannot receive copyright protections. Various cases now headed towards Supreme Court where I’m hoping boundary will be further defined at the highest level, but a human must be the author. There have been few instances where copyright office now is granting compilation registration for AI assisted works. It’s a thin copyright but better than none. I’m not a patent attorney, but from what I hear from my peers is that there is a shift in asset value. So for example, AI generated software may not be protectable by copyright or a co-invented hardware may face patent challenges.
So AI is a powerful instrument in the orchestra. The asset value may still be protected by way of trade secrets and contract terms rather than a registered IP, but humans still have to conduct this orchestra. Otherwise, we cannot actually find a meaningful collaboration. Future, I cannot say as of now, but as of now, without question, human plus machine.

Tom Stoneman:
So with all that going on, is the value… Because a big value of many companies is, well, patents, of course, but trademarks and copyrights, is the value of those assets going down?

Vintee Mishra:
When it comes to trademarks, I don’t think current LLM based models have appropriate trademark search database to find and compare what is already existing because of the race to market the products, AI product companies. I don’t know how much and to what extent they are going to do those searches and therefore a lot of litigation’s coming on that front as well, which is translating AI uncertainty into business solutions. And I don’t want to name the giants, but there have been instances in trademark litigation space where they adopted product names which were already existing in the market and had to enter into trademark litigation. They got narrow rights, not full trademark protection, right? So there have been an issue. There is no asset value down grade as such, but going forward, AI product companies will have to be mindful of working within the established legal protection guardrails because laws are coming up, trying to catch up.
And though they are slow in catching up compared to the innovation that is happening at a rapid pace, it will still have value attached to it. Will it be there like the way it is today is questionable, but there will definitely be not asset devaluation in the current landscape. 10 years later, maybe IP may not be as relevant, but what has happened in the past can also not be taken away in its entirety. There will be new valuation models to assess the value of the IP company or a person has generated in the past or in future.

Tom Stoneman:
So clearly AI presents some incredible new opportunities in the field of law, but it also presents some complications. AI tools can bring down the turnaround time on things like first pass contract reviews from weeks to hours, but AI’s output is far from perfect. It can have a tendency to hallucinate, so everything it does still requires human review and sign off. And while generative AI has been a huge boom to efficiency in some sectors, at the moment, AI output cannot be copyrighted, and the rules around what constitutes human assisted or purely AI generated are murky. So for organizations, it’s really about making sure you’re staying on top of the law and also putting in place rigorous systems of quality control. And you can see how for professionals like Vintee, navigating this new world of AI legal tools could be exciting but also stressful and anxiety-inducing. Thankfully, when Vintee’s feeling overwhelmed, she has a furry family member to turn to.

Vintee Mishra:
We have a goldendoodle. He just turned one in January. His name is Alpha. He’s definitely leader of the pack at home. Our life revolves around him. My son considers him as his brother, his best friend, and together they both just bring so much energy in the home. There is never a dull moment. Anybody who’s a dog lover would know the kind of energy and positivity they bring to your life, to your family, to your homes. He gives me a complete mental context switch when I’m deep into something and something challenging. So yeah, my breaks are usually around that. And then some family time, some cup of chai or a tea. Being an Indian, I think it’s just dissolved in my blood and chai is such a staple for us that it instantly energizes us. So, yeah.

Tom Stoneman:
As an attorney and focused on technology, SaaS and cloud agreements, software licensing, data privacy, how do you translate all of the uncertainty and legal risks associated with AI implementations into practical business-ready solutions?

Vintee Mishra:
Yeah. So as a contracts manager, what I focus on in contract negotiations first and foremost to understand where is the data going, what underlying model is the vendor actually using. It is important for me to understand that we have built human review and creative judgment to AI workflows, not just for quality control but also as a step to preserve IP rights like we were talking earlier. I have to ensure that I diagnose the problem before choosing the solution because legal tech fails when teams buy before agreeing on the problem. I have to definitely understand what IP indemnification protections exist, how is liability allocated if AI output causes harm. The goal is to build the right contractual architecture around the tool and not block the tool or the deal entirely. Because it is an emerging landscape technology, we have to ensure that we are not entering into multi-year contracts with auto-renewals. We have to ensure that we are evaluating and reviewing those contracts against our playbooks and templates because they drive adoption, which is the real measure of success.

Tom Stoneman:
In the marketing organization where I sit, there’s a lot of enthusiasm or desire to try to get these tools implemented without really thinking about the legal implications. And I don’t want to think about them most of the time because I just want to get the outcome, right? So with all these things, where do you see all of this going in the next six months to two years? I mean, what kind of changes do you see happening here?

Vintee Mishra:
From the demos I have seen in the past, what I’m hearing from several peers, and I mentioned contract nerds reports earlier, it looks like 90% of AI redlining tool features are identical and they need to be consolidated eventually to narrow down and achieve service quality and adoption support. Like we briefly touched upon on the fact that there are thousands of tools competing in this space right now, that pace is unsustainable in my opinion. Enterprises will have to get more selective, so the procurement tightens up and legal and compliance get earlier input. They need to be involved from the beginning. It’s already a billion dollar market as of, I think, mid-off last year. I’m sure it is multiplying at a rapid pace. We’ll have to eventually consolidate tools, legal frameworks, and market players. That’s what I think will happen in the next one to two years.

Tom Stoneman:
So what’s one thing out of all of this that you’ve learned either about technology or the law that you think more enterprise leaders should understand?

Vintee Mishra:
We have to understand that what will be the most consequential IP development in years, we’ll have to see what kind of laws are coming up, signaling global momentum towards AI identity protection. They’ll have to understand that there is no substitute or replacement for genuine human connections, which requires presence, empathy, and vulnerability eventually.

Tom Stoneman:
Being someone who sees the legal ramifications of AI pretty much every day, I imagine, is there any part of your life where you prefer just a human touch over AI or any part of your life you think that should be kept AI free?

Vintee Mishra:
My family and my friends, those relationships for sure. We definitely require genuine human connection, requiring presence, empathy, and vulnerability, human accountability for high stakes decisions because we just cannot completely rely on machines to make that happen. We must incentivize human creativity. We must incentivize people and find ways for them to be involved in the community, in their financial, mental, physical wellbeing when AI becomes more and more robust and more and more capable of taking care of different areas of our life. So yeah, human judgment, human accountability, human relationships must remain at the center.

Tom Stoneman:
Can you imagine something that AI will be able to do that’s never been done before?

Vintee Mishra:
I think I can speak from the contract point of view. There will be more and more real time dynamic contract negotiation. I think more and more AI systems will start negotiating from both the sites simultaneously eventually. There’s already a foundation for legal tools to help in executing the agreements, but AI playbooks that can learn from executor agreements will have better probability of helping lawyers to spend less time on mechanical drafting, more on strategic and relational work. And week long processes will eventually then will get compressed into hours. It’s already happening. It will happen at a much faster rate. The human-centered work is still where the law says the value lives. It was affirmed in some major case laws earlier last year as well. So yeah, it will do a lot of things that have never been done before. Future legal teams will be cross-functional, they’ll be agile, they’ll be tech native, and we’ll have to start building towards that model now.

Tom Stoneman:
Thank you for being on. We really appreciate your taking the time. I know how busy you are, and we’re just really thrilled to have you on the episode.

Vintee Mishra:
The pleasure was entirely mine.

Tom Stoneman:
AI agents negotiating a contract entirely autonomously, it feels like science fiction, but at the pace we’ve seen AI evolve, it’s probably closer than you think, and that’s part of the problem here. These tools are evolving so quickly that it’s tough for organizations to keep pace. But as Vintee said, trying to regulate these things only after something’s gone wrong, like a missed term in a contract or a fabricated piece of case law, well, that’s probably going to lead to trouble. That’s why it’s essential to make sure you understand the problem you’re solving before selecting the right tool and ensuring the robust guardrails are in place to prevent misuse. And I know sometimes people feel that these rules are stifling or restrictive, but it’s these safeguards that ensure a smooth rollout for these AI tools and really help unlock their potential to drive efficiency and profits. Thank you for listening to The Intelligent Enterprise, a podcast where we get inside big ideas by getting outside of them. I’ve been your host, Tom Stoneman. Please remember to follow the podcast and leave comments or reviews wherever you get your shows. See you next time.