Log in to watch

Log in or create a free account to watch this video.

Log in
AI Summit Bletchley Park 2026
Share

Fireside With Lord Tim Clement-Jones and Gene Kim

Lord Tim Clement-Jones CBELT
Liberal Democrat Lords Spokesperson for AI and Digital; Co-Chair, APPG on AI, House of Lords
Gene KimGK
Vibe Coder, Author & Researcher, IT Revolution

Gene Kim sits down with Lord Tim Clement-Jones, the Liberal Democrat Lords spokesperson for AI and digital and co-chair of the All-Party Parliamentary Group on AI, to talk about how AI is changing the legal profession. They discuss the future of the billable hour, how junior lawyers will learn when AI does routine work, what boards should ask about AI, keeping the human in command, managing AI risk through standards, and where UK regulation stands.

Chapters

Full transcript

The complete talk, organized by section.

Host Intro (Gene Kim)

For our last session before lunch, I have the privilege of sitting down with Lord Tim Clement-Jones. Lord Clement-Jones is the Liberal Democrat Lords Spokesperson for AI and Digital, and Co-Chair of the All-Party Parliamentary Group on AI, which he co-founded in 2017. He chaired the first House of Lords Select Committee on Artificial Intelligence, and he wrote the fabulous and serious book, Living with the Algorithm: Servant or Master? He is a lawyer by training and was London Managing Partner at DLA Piper, where he still advises on AI policy and innovation.

So we were introduced to him by Baroness Kate Rock again, who served with him on that select committee. And so when we talked this summer, I asked him how AI is changing the legal profession. I wrote down so many notes that I thought would be so relevant to this group of leaders, even though most of us are not lawyers. I asked him if he'd be willing to share his advice and perspectives, especially the advice he would give to managing partners.

So please join me in welcoming Lord Clement-Jones.

Gene Kim

Hi, Tim.

Lord Clement-Jones, you wrote a wonderful book. And I thought it was a very serious book for serious people, Living with the Algorithm. It's clear that you're enthused about the promise of AI, but what application of AI have you seen that fills you with most awe and wonder?

Lord Tim Clement-Jones

Well, the last session was pretty good, wasn't it? But I reckon it's in the health area, and I think what Demis Hassabis and his team have done at DeepMind with AlphaFold, and just getting the architecture of molecules that they have, and the possibility of drug discovery being speeded up, cancer vaccines, all of that.

I saw some of the results of that only in the last week. A team from Oxford, they're piggybacking on some of that technology, and that is going to be game-changing. Personalized medicine, mRNA-type technology being applied to the individual. This is going to be absolutely game-changing.

Now, I don't want to be Jeff Bezos and live for 1,000 years, or indeed President Xi, who I think wants to do likewise, but I do think they're going to help us remain in good health for much longer.

Gene Kim

Fantastic. And by the way, congratulations on Sir Demis Hassabis' Nobel Prize on that work.

Lord Tim Clement-Jones

Yeah.

Gene Kim

So when we talked, I shared a story about a friend of mine. He's a managing partner at a mid-sized law firm in the States, and he spent years easing one generation of equity partners out of the firm because of the drag they're creating on the active partners. And then he had to deal with a generation of associates that didn't seem to have the ambition to become partners.

And when I asked him about what he's going to do about AI reshaping his industry, he laughed and said, "I think I'll leave that to the next generation because I'm retiring. I've done my part." I told you that story, and you had a very surprising reaction because instead of laughing, you said, "That's a real cop-out."

Most people in this room aren't lawyers, but I thought the advice that you would give to managing partners seemed so relevant. So to my friend who's retiring, what is he choosing not to see?

And you had a story about how much law is changing, that instead of coming for advice, they actually come with a contract with ChatGPT outputs. Tell us about how you see the firm and the field changing.

Lord Tim Clement-Jones

Well, just to confirm, I'm probably older than that guy.

Gene Kim

Oh, yeah.

Lord Tim Clement-Jones

And quite honestly, I absolutely think that it's a cop-out. I don't think my generation, even my generation, the baby boomers, if you like, should be copping out at this stage. We've benefited hugely from the changes in the economy during our lifetimes, and I think now we need to show some leadership in these areas.

And the legal profession—and this isn't a session designed to show everybody how their legal bills are going to go down—but it is a fact that the productivity of use of AI, things like Harvey and Legora and so on, which are now very standardized tools, can really help in-house lawyers navigate their way, then say to their relationship partner in the external law firm, "Look, what do you think? Have I come up with the right solution?" So this, in terms of private practice, is absolutely game-changing.

In the old days—when I say the old days, probably about five minutes ago—many of these law firms still charge on an hourly rate basis. It was a question of almost weighing the file, saying, "How many hours have I spent on this matter?" Or, "That's it, I multiply my hourly rate by the number of hours, and Bob's your uncle."

Now, when you're a relatively senior partner and you're using AI to do your research for you and so on, you can't do that any longer. You've got to consider yourself part of the mainstream. When did McKinsey ever charge by the hour, for heaven's sake? When do some of those major consultancies charge by the hour? They charge by project. They charge by the expertise that they have.

And I used to run a government affairs practice within a law firm, and I never charged by the hour. I thought, well, look, if I can answer somebody's political problem in quarter of an hour, does that mean I charge them for quarter of an hour's time after a lifetime of experience in a particular form of politics? No. You've got to basically look at the quality of the service you're offering, and that's what consultancies do, and lawyers are going to have to get used to this.

And the quality of the advice, the relationship, the trust that their clients have in them is going to be absolutely crucial, not simply every year thinking, "Oh, I'm going to put my charge up from 900 pounds an hour, 1,000 pounds an hour to 1,500 or whatever it is, and then I'm going to charge up that hourly rate." That is just not the way forward.

Gene Kim

And you had another observation about how the work with clients has changed, not just law, but also medicine, where patients are coming in armed with lots of advice from AI. Tell us more about that.

Lord Tim Clement-Jones

Well, this is what I said at the beginning. Basically, in-house lawyers now can get an awful lot of what they need from these AI tools. And things like Harvey you can buy in, you can buy in Legora, which is the European version, and you're halfway there. And you're using fewer people, your productivity is higher, and so on.

There is another very crucial aspect because no longer does an in-house lawyer necessarily have to deal with the junior lawyers. You may find there are no junior lawyers knocking around there. And this is one of the real problems that actually law firms are having to address now. How do they train up younger people to take over from them in the future?

If they simply carry on at the moment, which is having entry-level jobs eliminated to quite a high degree, they're going to find that they actually don't have a law firm left in the future when this senior partner may be retiring, but these junior partners are going to be retiring in 10 years, 15 years. So they've got to think about things like that.

You've got to think about, do you train using virtual reality? Do you have kind of war gaming, red teaming? Those kinds of training experiences are going to be really important for some of the younger lawyers who can't just simply expect to tag along with the senior partners because the senior partners are using AI to answer quite a lot of their technical issues.

Gene Kim

Right. Maybe you can see why I found talking with Lord Clement-Jones so captivating. I had shared with you that one of the more interesting airplane flights I've ever been on was sitting next to a managing partner of Perkins Coie. And he explained to me that when someone comes out of law school, they're not competent. So they'll spend four or five years training this person to be competent. But then for the next four or five years, they're still losing money for the firm. So you have to train them to build a book of business so that they can actually add value to the firm.

So what is my friend, the managing partner who is choosing to retire early—what advice would you give these generational managing partners about what law firms should be shaped like going into the future?

Lord Tim Clement-Jones

Well, I think what he's got to do is think about how he trains up his people. But it isn't technical. It's about how you develop judgment, how you understand what's going on in a negotiation, for instance. Where do you think the weaknesses are? Psychologically, do you think they're beginning to crack? There are ways and means of doing this.

And let's face it, that reminds me, the Northrop Grumman presentation reminds me that an awful lot of what the army does, what an awful lot of the services do, is train. They have exercises. Well, that's what we need to have. It doesn't have to be real-life action.

If you're training up younger lawyers, it may be that what we have to do is the firm has to make enough money to be able to pay for exercises, for red teaming, as I say. So you do develop that judgment. It can be simulated, but nevertheless, you're building up expertise.

And that's got to be thought about now because if anybody's got any offspring who are looking to join the legal profession, well, they're all finding it pretty difficult at the moment.

Gene Kim

And so you had, I think, unkind words about the senior partners who are merely fee earners. And specifically, you said when you retain a white-shoe legal firm, what is it that you're actually buying? Could you teach us about what would you teach senior partners about what the real value proposition of the law firm really is?

Lord Tim Clement-Jones

Well, it's all about trust, isn't it? Why do people stay with particular law firms or particular individuals in law firms? It isn't really a brand in many ways. You hope that as you pass from one partner to another over the years, that the brand delivers a certain quality of advice.

But actually, it's about the individual. It's about whether you trust them, the kind of relationship you've got, whether they're going to be utterly frank with you about the issues, whether they've got a sense of humor, for heaven's sake. Sometimes it's pretty important. And I tell you, I've fired quite a lot of lawyers in my time for not having a sense of humor.

So I like the relationship that you have where you can be perfectly frank. But it is. It's something that's almost in the water. And that's what makes a good lawyer, a trusted lawyer in a major law firm, really quite unique. And developing those qualities in the younger lawyers is a real challenge.

And so your senior partner retiring really shouldn't be copping out.

Gene Kim

I can't wait to show this video to my friend.

You painted another dynamic that I thought was riveting, which is that as the economics of research and legal work change, there's a potential risk of a race to the bottom where we've commoditized that work and now we're all competing on price. Talk more about that, and what is the counter to that?

Lord Tim Clement-Jones

Well, this is always the fear of lawyers, that everything gets commoditized because once you've developed a way of doing things, that becomes standard. The really important thing is to keep moving, is to be creative.

We talk about the creative industries, but actually, some of the best lawyers, best accountants, best consultants are creative in their own right. So we mustn't just simply think of the theater and film and television and so on as being creative. Professionals can be highly creative as well, thinking of new ways of tackling things.

Some of the most creative people I know are in an area of law that you and I would think was dull as ditchwater, which is the reconstruction side of things, where companies are about to go belly up, or they're going to do a pre-pack, or they need to find ways of getting rid of the debt that they've accumulated, or there are some real issues about whether or not they're viable going forward.

Well, some of the most creative lawyers I know are the reconstruction lawyers, and it's highly competitive, and they are highly valued by the people that they work for. And quite often, the firms that employ them are specialists themselves in reconstructions and refinancing. And I call that creative.

Gene Kim

Suddenly I feel like a career change. Law sounds very fun to me right now.

During our conversation, you had mentioned the notion of the board having to be responsible for strategy. And after our conversation, I wrote down this. It evoked another conversation that I had with a co-author of mine, Dr. Steven Spear, who wrote the most famous Harvard Business Review article of all time called "Decoding the DNA of the Toyota Production System."

But I recently heard him earlier this year say, "Leaders at the top are too few, too slow, too far from the work," and in his words, "too stupid to solve problems." And so really, we need to rely more on people closer to the work, the front line.

So when you say AI strategy should be owned by the board, how is it possible that the board's opinion is of value? What can we do? What useful advice can boards give to the managing partner in a world where the answer to most questions is nobody knows?

Lord Tim Clement-Jones

Well, I think what you've got to do as a board, and I think this is true whether or not you're on the board of a law firm or a PLC or a startup or a scale-up, whatever it may be, you incorporate the values of the business.

Now, I think you need to have a pretty clear idea of what your values are as a business, what your business purpose is, and how you're going to take that forward. And so even if you are not completely au fait with every aspect of AI, for instance, I think you should educate yourself in any event. I think your retiring partners would spend some time reading a few books about AI, probably, and that would be an advantage.

But given that we're talking about the future and we're talking about current board members, I think it behoves those who are now on boards to get to grips with this and try and understand it. And then say, well, look, if we're going to adopt AI within our business, by what values are we going to judge the kind of adoption that we need to have? Are we talking about, for instance, augmenting the work of our employees, or are we substituting for them? Basic things like that.

How transparent is the use of AI in our business going to be? What's the purpose of it? And you go through a series—I wrote a chapter for an Institute of Directors handbook on AI governance, and there are about 20 questions that you can ask yourself about AI and how you're going to use it going forward.

And agentic AI, of course, now raises a whole bunch of questions. Not quite as serious as the ones we've just seen, because it's rather different working in business than in defense. But how much decision power, in a sense, do you allocate to your agent versus the human in the loop or the human in command, I think, is extremely important. So that is a kind of value system that you need to build in.

I'm a great believer in trying to have the human in command so that you have—it could be the CTO, but actually, in my view, the CEO needs to know what's happening. And then that is part of the governance within the business.

Now, it's really important that boards get the governance right currently because we have no particular regulatory framework that dictates how people need to adopt AI or indeed what AI developers develop. So that consciousness of governance and the values that should inform it is really, really important.

Gene Kim

Yeah. And so clearly, an area of passion for you has been not just AI as a topic, but also the government's role in it. You talk about oversight, not just at the government level, but other sort of mechanisms like AI governance. Tell us what you think is most important and what are the risks of getting that wrong.

Lord Tim Clement-Jones

Well, you use the word risk there. I think that what you need to be very conscious of, and you need to build in tools to help you, is assessing risk up front of the actual adoption of AI. I think if you're a developer, you should be assessing the risk itself up front. But we're talking about adopters here, really.

So that means, therefore, you need an impact assessment, an algorithmic impact assessment up front. You need to know the impact on your employees, on your customers, your consumers, society more broadly.

And then the consequences of that is, well, you see if it's high risk, certain things follow. You need to make sure that your transparency requirements are being adopted, that you're continuously monitoring how the AI is operating.

And there's a series of actual international standards that you can adopt. 42001, ISO 42001, which is a management standard, can be adopted for how you actually use AI in your business. And there are many others, which are pretty standard, and there's a US standard risk framework that can be adopted as well.

So there's no shortage of governance tools that you can use when you're thinking about this. You just need to be conscious that you have to do it.

Gene Kim

Yeah. By the way, I'm not sure if—Ned didn't disclose this when we talked, but for about 10 years, I was a card-carrying member of the Institute of Internal Auditors, and so the whole notion of boiling down to risk is actually a heck of an insight, to say that business risk is still business risk.

So three years ago to this date, the first AI Safety Summit was held in these grounds, and the Bletchley Declaration was signed here. You chaired the first Lords Select Committee on AI. You've co-founded the APPG. So of all the things you're working on as a legislator right now, what is the one thing that you are most passionate about getting right, and what should leaders in this room do about it before the law makes them?

Lord Tim Clement-Jones

Well, we tend to flit from tree to tree. First of all, a shout-out to Baroness Rock, who was on my select committee a few moons ago, and also a shout-out to Bletchley because this is where my parents met. So they were codebreakers during the Second War, and they met here.

And of course, Bletchley is positioned very close to the railway station. That was the whole point, so that the codebreakers and the employees here could commute to Bletchley Park. And I've been a member, if you like, of the community of Bletchley Park ever since the idea of making a museum here was started, and I've got a commemorative brick for my parents just out there. So that's by the way.

But Bletchley does represent something. It represents the benefits of technology, and I think what we have to make sure is that we get those benefits and the opportunities, whilst making sure that we mitigate the risks. And risk mitigation is one of my key things.

And when I say we flit from tree to tree, I mean, we might have the Data Use and Access Bill, now Act, one year. We might have the Digital Markets Bill, now the Digital Markets, Competition and Consumers Act. And we're just, today, saying goodbye to the Cyber Security and Resilience Bill, which will become an Act very soon.

And we've been arguing that agentic AI is creating all kinds of real issues in cybersecurity, and we're not getting that right in legislation. We're not being specific enough. People say, "Oh, well, you can't tell. The risks change all the time." But you can deal with what's in front of you.

And at the moment, we feel that the whole experience of the last two or three months, where we've seen agentic AI tools escaping from sandboxes. We've seen our own AI Security Institute fail, really. It gained access to the internet, which was a bit crazy.

Gene Kim

Yes. What the hell?

Lord Tim Clement-Jones

They saw some real problems arising. We don't feel that the legislation currently gives enough power to government to say, "Switch this off at the data center point," or even to understand enough about what's going on.

We've got notification, but we've got no power, for instance, in the AI Security Institute, to say, "We must inspect your tool." It's all voluntary, and they have no power to say, "Look, we don't think that this is a safe AI model. You must adhere to these standards." And they have no power to do that.

So there's no regulatory framework. It's all about information and notification and so on, which, going forward, I don't think is enough.

Gene Kim

So one of the traditions we have is that every speaker ends with a slide that says, "Here's the help I'm looking for." So among this community of technology and business leaders who are very switched on in terms of applying AI to their organization strategy and operations, what help would you be looking for?

Lord Tim Clement-Jones

Well, I think it's perspective. I spend a lot of time talking, and that's why I do things like this. I think it's really important not just to sort of pontificate inside our gilded chamber in Westminster. I think it's really important to talk to business of all kinds about the challenges that they have. What kind of regulation? Do they need regulation? Is governance sufficient? All the questions. And then what kind of governance?

And one of the really important things that I found of late is the sovereign AI agenda. Because one of the great frustrations is if all we do is, for instance, in government, buy tools from US hyperscalers. Love them though I may do, I use Claude on a daily basis. But the government should be helping startups and scale-ups develop their products, and they should be encouraging the use of open-source, open-weight AI tools to make sure that we are developing our own resilience in this area.

And at the moment, I'm not satisfied. We talk a lot about the government sovereign procurements and things of that sort, but we're not there yet. So I think we have to really gear up.

And the more I hear from business, the less geared up I think our procurement agencies in government—things like G-Cloud, for instance—not fit for purpose. So I think that kind of perspective and contact with business, UK business in particular, is really important.

And I never think that, if you like, AI sovereignty is independence. Of course, it isn't, but it is resilience. Say, for instance, we had President Trump decide that nobody was going to get Mythos for a few days. Well, if that had lasted longer and people had relied on Mythos, Anthropic's Mythos, then maybe that would've been a serious situation. Well, that may occur in the future.

The CLOUD Act may be invoked by the US administration at some stage, which would mean that there was a greater control over data held abroad, and much of the data that we house with Microsoft is held abroad. So we're not really gripping some of this, particularly in cloud services.

Gene Kim

And so people also interested in this topic will be truly interested in the talk from Babcock and Coastline tomorrow. So, fantastic.

Ladies and gentlemen, Lord Tim Clement-Jones. Thank you so much.

Lord Tim Clement-Jones

Thanks.

One membership. The IT Revolution library.

IT Revolution's books, the video library, the courses, and the MCP. The papers stay free.

Coming January 2027