Leading AI transformation at Brækhus through people and culture - insights from Atle Carlsen

As managing partner of Brækhus, that belief shapes how the firm approaches one of the biggest changes the legal industry has seen in a long time. Because while AI is moving fast, the real question is not how powerful the tools are. It is whether firms are able to turn that potential into something real in everyday work.

That is where many firms get stuck. The technology may be impressive, but if people do not understand it, trust it, or know how to use it properly, the effect stays limited. For Atle, that is why AI cannot be treated as a side initiative or a purely technical project. It has to be part of the firm’s strategy, culture, and operating model.

At Brækhus, that is exactly how they approach it.

AI is not something the firm is experimenting with on the margins. It is a strategic priority. Not just for a quarter, and not just because the market is talking about it, but because they believe it will fundamentally change how legal work is done.

“This is our main focus this year,” he says. “And probably next year as well.”

That view comes from a broader reading of where the market is heading. Clients are changing. Expectations are changing. Technology is changing. New players are entering the market, ownership models are evolving, and the pressure on law firms to work in smarter and more efficient ways is increasing. At the same time, clients themselves are becoming more capable. Some already use AI on their own. Some come prepared in ways they would not have done a few years ago. Many expect their advisers to use the best tools available - and they expect the gains from that to benefit them too.

Atle is clear that this is reasonable.

“The gains from AI are something clients will expect to benefit from,” he says.

That creates pressure on the traditional law firm model. Work that used to take hours can now be done faster. Simpler tasks may no longer need a law firm at all. Clients will handle some of them themselves. Which means firms will need to move more clearly towards the work that is hardest to replace: complex legal issues, judgement, strategy, and high-value advice.

That shift is already underway.

At Brækhus, one of the first things they realised was that choosing the right AI partner was not only about the platform itself. It was just as much about whether that partner could help the firm build internal capability.

When evaluating providers, they looked at two things. The first was the quality of the technology. The second was the provider’s ability to bring competence into the firm.

That second part mattered most.

“We have always believed that the bottleneck is the people,” Atle says.

That is why training and adoption became central from the start. The decision was not just about buying access to a tool. It was about finding a partner who could help build competence across the organisation in a practical and credible way.

At Brækhus, that effort has become highly structured. Atle describes the firm’s AI implementation through six steps. The important thing is not just the model itself, but what it says about how seriously they take the work of making AI part of the business.

The first step is leadership. People need clarity on what matters, what is expected, and where the firm is heading. At Brækhus, AI has been made a clear priority. Training is mandatory. Participation is not optional. That, in Atle’s view, is a leadership responsibility.

“You can’t ask people to do something if you’re not doing it yourself.”

The second step is security. Before people can work confidently with AI, they need clear guardrails. The firm has established a code of conduct for how AI should and should not be used, and people are trained in that as part of the rollout.

The third step is training itself, and here the firm works on several levels at once. There is a formal level, where everyone goes through compulsory training. Then there are internal AI champions who run more specific, theme-based sessions. On top of that, there is an informal layer: regular forums where people share best practice, ask questions, and learn from one another.

That last part matters more than it might seem.

Atle’s ambition is not simply to make people use AI. It is to help build what he calls a learning organisation around AI. That means creating an environment where people feel safe enough to ask questions, admit uncertainty, and share mistakes.

“If someone makes a mistake, we want them to share it,” he says. “So others don’t make the same mistake.”

That kind of culture does not happen automatically. It has to be modelled. It has to be encouraged. And it has to be built into the routines of the firm. At Brækhus, AI is a standing item in department meetings. There are regular arenas for discussion and sharing. And as a leader, Atle believes he has to go first - including when it comes to talking openly about his own mistakes.

Trust, in other words, is not something you announce. It is something you build.

That becomes especially important in a field like law, where people are trained to question, verify, and think critically. AI cannot remove that. If anything, it sharpens the need for it. One of the reasons Brekhus puts so much emphasis on learning and culture is precisely because responsible use depends on people understanding what the technology does well, where the risks are, and how to work with it critically.

But training and trust are still not enough on their own.

The fourth step in Brækhus’ model is what Atle sees as the most important one: getting AI into the core processes of the firm. Because firms can have high levels of AI adoption without seeing much measurable gain. People might use the tools now and then, but unless AI becomes part of how work is actually done, the real value stays limited.

“If you don’t work on core processes, you won’t get the gains,” he says.

That is why each department at Brækhus is mapping its core processes and identifying where AI should be used systematically. This is not always easy in legal work, where the idea of a “core process” is not necessarily intuitive. But Atle believes it is essential. Once those processes are visible, the firm can start improving them in a deliberate way.

The fifth step is incentives. Brækhus, has made AI engagement something that is rewarded. Employees can be recognised for contributing positively to the firm’s development through AI. More unusually, the firm has also built this into its partner model. Partners who actively use AI can benefit directly. It is a concrete signal that this matters.

And the sixth step is measurement. If a firm wants to know whether it is moving forward, it needs ways of tracking progress. So each department is working on KPIs that make it possible to see whether the effort is actually producing results.

This matters because, as Atle points out, a lot of companies currently have high adoption rates and very limited documented gains. The missing piece is often structure. Without process work and without measurement, AI remains promising rather than transformative.

That is also why Atle sees this as a source of competitive advantage. Firms that approach AI systematically - not just enthusiastically - are more likely to get somewhere meaningful with it.

Even so, he is realistic about where things stand today. The biggest gains so far are often found in tasks like document review, due diligence, and turning around reports faster. Those are real gains. But they are still only part of the picture. The larger opportunity lies in changing how the firm works at a deeper level.

And beyond internal efficiency, the relationship with clients is also changing. AI will allow clients to do more themselves. Some of the more routine work that once naturally sat with law firms will gradually disappear or be absorbed elsewhere. That does not mean law firms become less important. It means they need to become more focused on the work that clients cannot easily replicate on their own.

That includes judgement. Strategic advice. Complex legal interpretation. The human side of difficult decisions.

In that sense, AI does not reduce the importance of lawyers. It raises the bar for what clients expect from them.

For Atle, this is why AI strategy should never sit separately from business strategy. It has to be tied to where the firm wants to go, how it wants to compete, and what kind of value it wants to create for clients.

At Brækhus,, that means using AI not only to improve speed and quality, but also to strengthen the firm’s position in the market. It is about becoming better, faster, and more relevant at the same time.

What stands out most in the way Atle talks about this is that he does not present AI as a silver bullet. He talks about it as an organisational challenge. A leadership challenge. A culture challenge. The technology matters, of course. But it is only one part of the equation.

The harder part is making change happen in a way that lasts. And that, more than anything, is what Brækhus seems to be trying to build: not just AI usage, but AI capability, not just access to a tool, but a firm that knows how to use it well.