For Christian, Head of Legal at Nordic Semiconductor, that captures what legal AI is starting to change. It’s helping a small in-house team serve a large organisation, to get to the heart of an issue quickly, respond with more confidence, and spend more of its time where judgment matters most.

At Nordic Semiconductor, the legal team is four people supporting a global company of around 1,500 employees. Between them, they cover everything from employment and corporate matters to IP, security, contracting, and the day-to-day legal questions that keep a global business moving.

“We are a small team and everyone needs to be able to do everything,” Christian says. “Whether it’s leasing the new coffee machine, procuring our supply for the year or negotiating with our key customers, it’s the smallest and the largest projects in the company.”

That breadth is part of the challenge. The other part is context. Nordic Semiconductor operates across hardware, software, and cloud services, in an industry shaped by technical complexity and a constantly evolving geopolitical landscape. For lawyers inside the company, the task is to interpret the law in a way that fits the reality of the business.

“We almost every day find ourselves as the least technical people in the room,” Christian says. “That can be quite intimidating sometimes. But the environment we are in makes us more receptive to technology and change, and seeing what it can do.”

A team surrounded by technology, ready to apply it to legal work

At Nordic Semiconductor, AI is not a future concept. The business builds semiconductor chips with AI accelerators, provides AI algorithm development tools, and uses AI in customer support to serve customers in multiple languages. That proximity to innovation shaped the legal team’s mindset early.

Christian says the first time he saw a general-purpose AI platform produce legal text, the significance was immediate. Being a generic tool, the output was not perfect, but it was close enough to show what was coming.

Like many legal teams, Nordic Semiconductor began by experimenting with tools such as ChatGPT. Those tools proved useful for exploring. But over time, it became clear that it was hard to reach a level of value consistently.

Legora was introduced through one of Nordic Semiconductor’s external law firms in Oslo, which had already deployed it across its own lawyers. For Christian and his team, the recommendation came with trust, and with confidence that a serious market evaluation had already been done.

“We get a lot of pitches from companies trying to sell their various AI services,” he says. “Having the law firm come in and tell us the evaluation they had done, that was a big enabler for us because we trusted their very thorough assessment.”

The product itself then made the difference.

“The first time seeing Legora, and how it was packaged in a manner that met us where we are, and in the systems we already use, it was immediately apparent that this was taking what we could do with AI to the next level.”

The impact showed up immediately

Although Nordic Semiconductor has only been using Legora for a short time, Christian says the shift was visible almost at once.

“I see it in almost every deliverable we give to the organisation,” he says. “I see the emails from the team are much sharper and much cleaner in language.”

The most obvious gains have come in contracting, where the team can work with controlled inputs, known clauses, and repeatable workflows. That makes it easier to accelerate drafting, issue spotting, and negotiation support.

“We have a couple of negotiations now that we typically would allocate two weeks of drafting on our end,” Christian says. “Whereas now, the team will come back and promise next-day delivery because they know that they can get so much help in getting 95% of the product done.”

That speed makes a difference because it also changes what the legal team can do with its time. For Christian, the treadmill example is one end of that shift: being able to get immediately to the core issues in a large contract and translate them for management. At the other end is the broader opportunity to free the team from high-frequency, lower-value work and spend more time on strategic support.

“A part of our aim here with optimising and reducing the time spent on high-frequency, low-value work is to be able to give ourselves space to figure out where we are going and understanding how the legal team can provide value in that direction,” he says.

Creating more room for strategic work

For a four-person legal team supporting a global company, the real constraint is not only workload. It is reach.

Nordic Semiconductor operates across the Americas, Europe, and APAC, while the legal team is based in Oslo. That means the team cannot rely on hallway conversations or informal proximity to understand what is happening across every part of the business. They have to create the space to be visible, proactive, and engaged before issues become urgent.

“We’re not going to meet everyone around the coffee machine and get a sense of what’s going on,” Christian says. “We need to set time aside to make sure that we are visible, that people are familiar with us, and that they can call on us for support that they maybe didn’t know we can provide.”

That is one of the clearest strategic advantages of AI for in-house teams: not simply handling the same work faster, but making it easier to redirect time and attention to the places where legal can have more impact.

Judgment matters more, not less

Christian is clear that none of this reduces the importance of legal judgment. If anything, it raises it. Earlier in his career, he took real satisfaction in drafting from scratch. He still values that craft, and the experience that came with it. But he no longer sees manual first drafting as the best use of a lawyer’s time.

“I’m glad I learned that trade,” he says. “And I can apply that experience when I’m now supervising the AI doing it.”

That idea of supervision comes up repeatedly in how he describes the team’s approach. AI can generate convincing language quickly. The responsibility lies in knowing when that output is sound, when it needs adjustment, and when it needs to be checked from the ground up.

“We still 100% need to rely on our professional judgments,” Christian says. “There is no replacing my responsibility as an individual practitioner when I take that work and submit it to management or to other stakeholders.”

That also means being transparent about confidence levels. Sometimes an AI-generated output is enough for a first steer. Sometimes it needs proper verification before anyone relies on it.

“Only yesterday I sent out something, quite a large summary of a huge document, and immediately stated that the topics were picked out by me, but I haven’t verified it.”

For Christian, trust is helped by verification tools inside the platform itself.

“We do have quite high confidence in the results AI gives us, but it also gives us all the tools that we can use to verify where it got that information from,” he says.

Building company knowledge into the system

One of the most important parts of Nordic Semiconductor’s adoption has been giving Legora more context about the business.

The team has been building topic-specific databases using internal and external materials: previous annual reports, governance rules, commentary, strategic documents, and company policies. The aim is to make specialist knowledge more accessible across the team.

“For corporate governance now, we’ve used previous annual reports, external commentary on our annual reports, the Norwegian governance rules, and we’ve also been feeding Legora our strategic documents,” Christian says.

That makes it easier to use AI not just as a drafting assistant, but as a practical layer over the team’s accumulated knowledge. It also helps reduce dependency on individual subject matter experts for every first-pass question.

He gives export control and sanctions as one example. Rather than routing every clause back to a specialist, the team can build playbooks and context so that standard issues can be triaged earlier, with escalation only where needed.

“That is a valuable thing that we will continue to explore,” he says, “providing company knowledge and policies to the entire team to apply without going to the subject matter expert, at least at first pass.”

Rethinking the role of external counsel

AI is also shaping how Nordic Semiconductor thinks about outside firms. Christian says the team is using external counsel cost as a reference point to help quantify the internal value created by AI-supported work. This is not about replacing law firms. External counsel still plays an important role in specialist advice, strategic input, and accountability.

“It’s not about us trying to reduce the importance and role of the external counsel,” he says. “It’s us trying to send a signal that we believe in this technology and that we will see returns on the investment we are putting in.”

But expectations are changing.

“We are starting to feel more and more impatient with how long legal work should take,” he says.

That does not mean turning everything into a race against the clock. Christian still values the quality and care that external counsel brings. But once an in-house team sees how quickly strong first drafts, summaries, and issue lists can be produced, it inevitably starts to reshape assumptions across the legal ecosystem.

He also sees a broader role for law firms as trusted guides in AI adoption itself.

“For us, they are a very trusted and valuable advisor on AI adoption,” he says.

What the next generation of in-house lawyers will need

Asked what he would look for in future hires, Christian points first to curiosity. At a company like Nordic Semiconductor, lawyers already need a genuine interest in technology. Now that extends naturally into how legal work itself is evolving.

“We would rank AI, not necessarily that you need to have high knowledge or be fluent in prompting and using tools, but a keen interest to see how you can apply your profession with those tools would be absolutely central,” he says.

That does not replace foundational legal skills. Strong judgment, drafting, and analysis remain essential. But he suggests that the modern in-house lawyer needs a broader profile: someone who can communicate clearly, understand the business, engage with technology, and still apply a rigorous legal lens.

“We would value strong legal judgment, but communication skills and an interest in learning about the business and the technology.”

For graduates, that curiosity may be especially valuable.

“I think graduates today who can show that they are interested in AI, they would almost be fast-tracked into most jobs.”

The starting line is closer than people think

For legal leaders still hesitant to engage with AI, Christian’s advice is simple.

“Sitting on the fence has never really been a good strategy when it comes to technology,” he says.

His point is not that every team needs a fully formed transformation plan before they begin. It is that the barrier to entry is lower than many assume.

“The starting line is much closer than many believe,” he says. “It is actually possible to start with just a simple chat, upload a few documents, and see how that works.”

That practicality helps explain why the shift has felt so real for his team. It has not been abstract, theoretical, or over-engineered. It has been immediate, useful, and energising.

“What surprised me the most was how immediately it impacted our conversations,” Christian says. “I was thinking that this would be me pushing AI on everyone in the team. But instead it was apparent for everyone that this is something they could use from day one.”

That may be the clearest signal of all. At Nordic Semiconductor, the story is not simply about working faster. It is about a legal team rethinking what deserves its time, where its expertise matters most, and how a small function can have greater reach across a complex global business.

And, as Christian puts it, there is one more thing worth noting: “I think it’s just plain fun as well. People are having fun with it.”