
Episode 15. Innovation and the role of AI in Police Work with Norwegian Police
Discover how Norwegian police are transforming investigative work through innovation and artificial intelligence.
This episode examines the integration of AI in police work and investigative interviewing, advances in crime scene investigations, and the critical importance of accountability when deploying AI tools.
The speakers discuss challenges in keeping pace with criminal innovation, the necessity for operational efficiency, and how international collaboration strengthens policing practices. They also address the cultural shift required within police organisations to embrace the innovation of AI in police work while navigating bureaucratic complexities. Join host Børge Hansen in conversation with three experts from Norway’s law enforcement community: Kjeld Hendrik Helland-Hansen and Oddvar Moldestad (Forensic Investigators, Western Police District) and Bente Skattør (Senior Advisor ICT and Innovation, Norwegian Police).
When Police Innovation Meets Reality: Inside Norway’s AI Revolution
At a recent crisis exercise with the Norwegian Navy, a forensic investigator walked through a simulated bomb scene speaking into a headset. No notebook. No frantic typing. Just his voice, capturing every detail as he moved through the chaos.
Twenty minutes later, he had a complete report.
“If I should have done this in the traditional manner,” he said, “I’d use at least two days, perhaps more.”
This is innovation in policing – not in some distant future, but happening right now in Norway’s Western Police District.
The Trust Question
Bente Skattør, Senior Advisor for ICT and Innovation at the Norwegian Police, is leading the charge to integrate AI into investigative work. She’s acutely aware of what’s at stake.
“If we don’t move faster, I think we might lose trust; if we are lagging behind the criminals, that will immediately hit the trust for the police.”
Bente Skattør
The numbers are staggering. Norwegian police conduct 150,000 investigative interviews each year – every one traditionally typed manually. Meanwhile, criminals have embraced AI for deepfake voices and sophisticated scams. In Norway, AI-enabled fraud has now surpassed drug trafficking as a criminal enterprise.
Small Steps, Big Impact of AI in Police Work
What makes the Norwegian approach different is the philosophy: small, fast experiments with real officers in real situations. Not waiting for perfect systems.
The results? AI interview transcription in 90 seconds. Crime scene documentation cut from days to minutes. But most importantly, the technology is designed with officers, not for them.
Forensic investigators Kjeld Henrik Helland-Hansen and Oddvar Moldestad have tested voice-to-text systems in actual crime scenes, refined the templates, and brought colleagues along for the journey.
“The ones that will benefit the most from it are the guys typing with one finger on their keyboards. They will really see the benefits.”
Kjeld Henrik Helland-Hansen
Accountability Built In
For all the talk of AI, accountability remains central. Every transcription is verified. Every AI output is reviewed. The technology accelerates documentation, but humans maintain control.
The real test comes in crisis exercises – four major exercises so far – where the team deploys their tools in realistic, high-pressure scenarios. They’ve proven that the technology works when it matters most, in what they call “the golden hour of investigation.”
Beyond Borders
The team has shared their work across Europe through Europol, in Brazil at international conferences, and with law enforcement agencies worldwide. They’ve earned a Europol Innovation Award and global recognition.
But the awards aren’t the point. Criminals don’t respect borders, so innovation can’t either.
“I think it’s counterproductive to sit in every country doing the same kind of innovation with just a small variance,” Kjeld Henrik explains. The Norwegian team operates without financial commitments that would restrict knowledge sharing – because when one police force becomes more effective, it raises the bar for criminal operations everywhere.
The Revolution
Innovation in policing isn’t a future promise. It’s happening now in police districts across Norway, driven by investigators who understand both the technology and the work it must serve.
The analog investigator who completed his crime scene report in 20 minutes didn’t become a tech expert overnight. He simply had tools that finally matched the way humans naturally work: by observing, speaking, and moving freely without being tethered to keyboards.
That’s the revolution – making technology fit the investigation, not forcing investigators to fit the technology.
“Innovation is a muscle that you have to train,” Bente says. In Norway, that training is already well underway.
Listen to the full conversation in Episode 15 of Beyond a Reasonable Doubt, where Børge Hansen talks with Bente Skattør, Kjeld Henrik Helland-Hansen, and Oddvar Moldestad about the real work of innovation in modern policing.
Episode Length: Approximately 59 minutes
Production: Davidhorn – Beyond a Reasonable Doubt Podcast
Host: Børge Hansen, CEO, Davidhorn
Equipped For Justice – Supporting ethical, human rights-compliant investigations worldwide
About the guests
Dr. Bente Skattør
Senior Advisor for ICT and Innovation at the Norwegian Police and project lead for AI in investigative interviews. She drives innovation initiatives processing over 150,000 police interviews annually, integrating artificial intelligence into investigative work while maintaining rigorous human oversight. Her work has earned a Europol Innovation Award, a National Digitisation Award nomination, and a Global Innovation Prize in Brazil. With extensive experience in project management across Nordic and global contexts, Bente specialises in the intersection of AI, big data, and law enforcement – focusing on investigative interviews, cybercrime, and creating innovation cultures in complex, high-risk environments.
Oddvar Moldestad
Police Superintendent with over 20 years of experience as a forensic investigator in the Western Police District of Norway. Over the past two and a half years, he has been actively involved in the AI4Interviews project, working to modernise and streamline forensic workflows through the use of smart technology and artificial intelligence.
Kjeld Hendrik Helland-Hansen
Police Superintendent working as a crime scene investigator with the Norwegian Police, specialising in forensic documentation and crime-scene methodology. He has a background in archaeology from NTNU and has worked for several years at the forensic unit in the Western Police District of Norway.
Kjeld has represented Norwegian policing internationally through the European Network of Forensic Science Institutes, as a delegate to EMFA under the ENFSI Scene of Crime Working Group. He has also served as the former head of the Norwegian Criminalistics Forum, an organisation for Norwegian crime scene investigators.
In recent years, his work has focused on innovation at the intersection of policing, technology, and research. He is a contributor to the AI4Interviews project, exploring how hands-free technology, speech-to-text, and artificial intelligence can improve documentation, situational awareness, and evidence quality in crime scene investigations.
Watch and listen also on YouTube and Apple Podcasts
Related products
Transcript
Host: Børge Hansen, CEO, Davidhorn
Guests: Kjeld Hendrik Helland-Hansen and Oddvar Moldestad (Forensic Investigators, Western Police District) and Bente Skattør (Senior Advisor ICT and Innovation, Norwegian Police).
BØRGE HANSEN: Today, we’re talking about innovation in policing and what it actually looks like. Criminals are already playing with AI tools. We see deepfake voices used to trick people, AI-written scams that feel uncomfortably personal. And at the same time, the amount of evidence is exploding. Every case means hours of video and audio, phone data, documents, and chat logs. The police aren’t just under pressure from offenders; they’re also buried under information. And still, many investigators are typing crime-scene notes by hand, replaying interviews over and over, and working in systems that don’t really talk to each other. So the question is, how do we keep up and how do we get ahead?
The good news is that innovation is already happening inside policing. And the team around this table has been recognised for that work with a Europol Innovation Award, a National Digitisation Award nomination, and a Global Innovation Prize in Brazil. I’m very happy to have you here. Bente Skattør, project lead and innovation lead for AI for interviews at the Oslo Police. Kjeld Henrik Helland-Hansen, forensic investigator in the West Police District and the wildcard. And Oddvar Moldestad, also a forensic investigator and a longtime CSI innovator in the West. In this conversation, we’ll look at what’s actually changing at the crime scene, in the interview room and in the courtroom. What works, what’s hard and what it looks like when innovation comes from within the police.
So let’s start there. Bente, you’re leading AI for interviews. When people ask you what is actually changing in policing right now, where do you begin?
BENTE SKATTØR: I begin with the police officers, the CSI folks, the team, because that is where you actually start. But first of all, thank you very much for having us here. We are really thrilled to be here to talk about what we actually burn for, and that is innovation in the police. As you so clearly mentioned, Børge, the criminals have already embraced AI.
And that is also where we are. We are really trying to embrace the technology, but of course, we have to do it in a responsible way. So what we’re doing now in the AI for interviews is actually trying to build upon the speech-to-text technology, AI analysis. So we are working with the police workers strongly and tightly because they have to take and use the tools.
So that is essential, I think. So what we are working on is actually, as you said, we are working with investigation, and AI for interviews is actually about—we started with investigative interviews. So that is where we have the starting point. But we have seen the huge potential of using speech-to-text in many areas. That’s why we are also here for the CSI crime scene investigation, the courtroom, and putting TVC on the streets from the patrols. So we have huge potential.
BØRGE HANSEN: What did you start with in investigative interviewing? I know Norway is really on the forefront there. Why is that and how does it work for you guys that you started out with something that the Norwegian police think they’re really good at already?
BENTE SKATTØR: Yeah, it was not a coincidence. Because I’m really proud also that Norway has a good methodology of the way we are doing investigative interviews. So we have sort of stolen with pride from the PEACE in the UK, but we have also cultivated our own methodology. But the point is that they are writing the investigative interviews manually.
And we saw also the potential for using Norwegian language models so they can focus on rather doing the summary or understanding what is in the interview instead of taking notes while they are doing the interview. And of course they are taking many. So in Norway, I know Norway is a small country, but anyhow we have approximately 150,000 each year interviews, whether they are inside or outside—150,000 interviews with victims, suspects or similar from the police.
So we have a huge and very exciting potential and we have come that far that I hope people can see or hear that we are on the right way.
BØRGE HANSEN: So what’s the impact that you’re seeking in AI for interviews? Obviously, you talk about the 150,000 interviews all across Norway. What’s the impact that you want to get out of those 150,000?
BENTE SKATTØR: As you said in the start, Børge, we are having so much data within investigation and we need to help the investigators to do their job so they can focus on what really matters within investigation. So we can remove the boring job. It’s very important to have a good summary, of course, that they still have to do. And they also still have to control what is coming out of these language models when we do the transcription. But the potential of letting them focus more on the core case—that is investigation. And the data is not going to stop. It’s going to flood more. It’s growing each day and I have no belief that it will be less, it will be much more.
BØRGE HANSEN: But you also work closely with, in Norway, we call PPS or police officers in the field, working in the field at the crime scene. How does that innovation work that you’re doing now? How does that impact the PPS situation?
BENTE SKATTØR: We are in a position now where we actually can ask them and they have answered a questionnaire, but they’re also doing interviews about what they feel. But today it’s going very fast from when you take the interview in the street until it’s transcribed. It takes approximately one and a half to two minutes.
BØRGE HANSEN: So from the interview performed to the transcription ready, a couple of minutes?
BENTE SKATTØR: Yes, absolutely. And also the data will be valid inside the police. But we can also enable that for the court, for operation central and the core team if we have a huge case going on. So they can see the data also, what’s coming from the street immediately. So we also built that inside.
BØRGE HANSEN: In our talks earlier, you often talk about innovation as a muscle that the organisation has to train. What does that look like in practice?
BENTE SKATTØR: Yes, we have big muscles. I always say that innovation is a muscle that you have to train in order to keep up the speed with the AI revolution which is ongoing. So you have to exercise on the turn of the technology. So that means we also have to train on different arenas. We have to do smaller proof of concept, very fast, fail fast, but also going in the right direction.
And then you have to work strategically. So when it comes to the smaller proofs of concept, doing the things to explore—then we have a really good example here, is about the crime scene investigation team. I’m really proud to have Oddvar and Kjeld Henrik on our team.
BØRGE HANSEN: Before we move on to the CSI team, we also said that police work and innovation have to go hand in hand. And what does that look like on a good day?
BENTE SKATTØR: When police officers say we are on the right way, this is good, it can be better. But when they say this is working and we are on the right—that’s giving me thrill. And I’ve lately been so lucky to hear that. And I also have seen that that makes us even stronger for striving for more innovation.
BØRGE HANSEN: If criminals innovate faster than the police, why should the public still trust the police to keep up?
BENTE SKATTØR: I think that is a huge question. First of all, if we don’t move faster, I think we might lose trust. Because if we are coming more and more or lacking behind the criminals, that will immediately hit the trust for the police. That’s kind of natural. And the police force is founded on trust. So if we don’t have the trust, it will be much harder to keep on going.
BØRGE HANSEN: So, Bente, one of the challenges we’ve spoken about earlier is that we feel that people challenge us that we need to know everything about everything before we can start using these systems. But your approach is much more sandboxing and point-to-point solutions.
BENTE SKATTØR: Yeah, and that is a huge… we can philosophise a lot and discuss whether we should have a general base model that can serve and suit everyone’s need, or you should go on point-to-point to actually deliver value. And when you are in that discussion, you can easily become paralysed. That is not a good thing because innovation has to happen fast. You have to respond to the needs and also to testing it out very fast on what works. So if you are going, and this is also what I’ve seen in several technology development in Norway, we sometimes think too much before we actually do something. So we have to look ahead, we have to do smaller point-to-point solutions, and if these smaller point-to-point solutions works, you can take the same piece of technology on the next thing, and then next thing. And that’s also the way we actually evolve the data.
BØRGE HANSEN: So working in police, one could imagine that you’re not only working with supportive people. There’s a lot of cynicism in police work because—well, the reality of living with crime and criminals every day. How do you take that cynicism and turn it into something progressive or positive?
BENTE SKATTØR: I think by meeting the police workers, I really do. Because when I speak about AI, the immediate question is, will the robot take my job? Of course we—I cannot, that is a question that is fair and you have to address it. So that is what you do. But that is also then why it’s so important to involve the police officers strongly and early. Because they see and hear what, because it’s hard to say, to think about what you need or what could help you. But once you see a proof of concept and you hear it in your own language, then you immediately see the help of how this will lighten your job. So that is by involving them, I think. And also admitting that, this is a sandbox, meaning that this is where we will test the new technology. But it doesn’t mean that everything will be working.
BØRGE HANSEN: As many people listening, working in investigative work or courts or prosecution knows, policing is a profession that’s caught between extremes. There’s a lot of rules and regulations and bureaucracy that stops work. At the same time, there’s an urge to do things in the right proper way because lives are at stake. How do you navigate that as innovators in a very traditional setting?
BENTE SKATTØR: Yeah, so this is actually what I’m burning for because I have worked now in 20 years or more within the police in ICT department, not operational. But I know that it is always saying that, no this is not allowed, or we cannot do that, or we have to check and so on. And I don’t say that we shouldn’t do that of course, we have to obey the law. But it is a huge difference saying what is possible under the law instead of saying, can we have the allowance? And I think that is a huge culture thing that we have to embrace in police as well. That we…
Of course, within the law, but what is the playground? So to have that change, you need the leaders that have that same value and will give you allowance to test things and that also okay to fail.
BØRGE HANSEN: For those of us who haven’t been to a crime scene, we have a lot of impressions from TV series. What is the job of working on a crime scene in 2024 or ’25?
ODDVAR MOLDESTAD: If I may say so, crime scene investigation as a profession is all about documenting. What you observe, you observe traces, you observe persons, you observe objects, whatever. And that is quite boring work in many ways because you have to be very thorough, you have to be structured. And with the Sandwich Project we now are part of, we see how we can do these tasks in a better way. What we want to spend our time on is that is very important for crime solving. But at the same time, we can let technology do the boring part, I mean the documentation part of it. So we can use more time with other observations.
BØRGE HANSEN: So can you just take us through a crime scene? We’re looking at a traditional crime scene, there’s been a breaking and entering, a house burglary. How does an investigator or a technical investigator work in that scene traditionally and how do they work now with technology that you guys have been working on?
ODDVAR MOLDESTAD: Just briefly, first of all we establish what has happened and what do we need to do to start our crime scene investigation. And that is actually looking around. And then you start doing your thorough investigation where we use, could be photography, all kind of methods to describe the scene. And you are very accurate, you make drawings, you measure, measure every object.
So we can try to recreate at a later point of time when you also done the whole investigation, make kind of recreation of what has happened. And yeah, that is a quite boring aspect of CSI work. So you have to be structured and write everything down and basically this is what we want technology to do for us because we see that could be a lot easier if you can have tools that can make this work easier, not only for us but of course also for the patrol officers. They have—they are first on the scene. They also need to do the same things we do, but of course they have other tasks as well. And yeah, we want to make this easier in many ways.
BØRGE HANSEN: Traditional CSI work is what? Pen and paper, dictaphone…
ODDVAR MOLDESTAD: Yeah, basically pen and paper. I think the last years we’ve had the ability to type into our tablets, but still you’re in this crime scene, you wear these gloves, you don’t want to contaminate anything, still you are forced to take up your phone or your computer or whatever and you have to type this in. So you don’t want to—if you find a suspected trace, you don’t want to take off your gloves and put on the gloves and do, redo this many times. So we saw quite early that voice-to-text could be the solution here. So that is what we started to do a couple of years ago when we started this project.
BØRGE HANSEN: So these days you are working in a crime scene with voice-to-text. What does it look like to you? What is your working situation like?
ODDVAR MOLDESTAD: I think you are perhaps more out in the field, and especially on this and testing this out in the—on the field. So perhaps you can elaborate more on that?
KJELD HENRIK HELLAND-HANSEN: Yeah, so initially when we saw the possibilities in this, I thought it would be revolutionary just to have something that can transcribe my voice into text. So I don’t have to stand there in the dark with gloves and having to write this. So that was just fabulous. And then when you also realise that this transcription actually systemises, from the voice to text, and systemises according to a template that we work with. So what we call people and what we call objects and what we call traces, it puts everything in a system. So basically we can go through the crime scene, describe everything we see and then get out a finished report and that is just fabulous.
I did some exercise with the Western police, and that was with a colleague who was also a CSI technician and also the patrol police. And one of my fellow colleagues, I would describe him as one of the more analog crime investigators in our office, but that’s fine, we’re different. He was using the headsets and just logging everything. It was counter-stabbing, it was an exercise with the Norwegian Navy where they had gone off two bombs and there were dead and wounded people—that was the scenario. And when we got to the place of the incident, then he started to log, describing everything and going along. And then this was transcribed and systemised. And we asked afterwards, so what do you think of this? And he just answered, yeah, honestly, I did this, I got this report now in 20 minutes.
And if I should have done this in the traditional manner, I’d use at least two days, perhaps more. So that was a very satisfying answer for us at least because we are tech nerds, I admit it. So we like to move ahead, we like to be in the front and testing things, but this has to have an impact on the whole level, not only as tech-savvy guys, but I think in this case, the ones that will benefit the most from it are the guys typing with one finger on their keyboards. They will really see the benefits of this.
BØRGE HANSEN: Crisis exercises, you guys said earlier that it gives you much more realistic data than synthetic test data. And as you describe, you can see the technology being used in practice with your own eyes. What’s the biggest surprises from these exercises? Good and bad.
BENTE SKATTØR: I might answer that because to having allowance to test out early technology is quite cumbersome. Sometimes you have to plan very thoroughly or maybe you have to work with synthetic data, but these kind of arenas is really fantastic to explore technology. And so we jumped very fast into the exercise because we had short notice—that’s also very cool. That was a level of, parts running again. But you can actually, because now we have a toolkit of tools that works pretty well. And we also gather data. So we are in the golden hour of investigation. So we use also the interview solution and also the CSI. But the most important was actually it not that it worked, but it was a good stage of exploring technology much better than I hoped for. So now next year we are actually planning to do more so we can expose and test our solution to the end users. And then we can build on something that is already working, but also have add-ons to explore because there’s no danger to having mistakes in such exercises. That’s what we are aiming for, to learn, build.
So I’m really thrilled about that. And one of the key questions that we have—we’ve also been participating in four exercises now. So we have data collected from these exercises. So when we have more power or GPUs or machine power in the police, it’s going to be really, really fun to run also AI analytics of this data. So yes, I’m really thrilled. And also what is… The people who work with this technology can be on stage to see how Kjeld Henrik and the bomb squad is using the technology. That’s essential also.
BØRGE HANSEN: Let’s zoom out a little bit internationally. These challenges aren’t unique to Norway. You’re watching and working with others around the world who are wrestling with the same problems. Who’s inspiring you internationally and what are they learning from your work and the Nordic approach?
KJELD HENRIK HELLAND-HANSEN: But I would also like to point out that we don’t have any financial commitment in many ways to this kind of thing. So we are eager to share. Now, criminals don’t care about borders. They don’t care if you’re in Norway or Denmark, or if you sit in one part of the world doing crimes in another part of the world. And I think it’s so important to share this kind of knowledge with our colleagues in law enforcement in other countries as well. Because I really believe that if we share this kind of knowledge, we will get something back. And I think it’s counterproductive to sit in every country doing the same kind of innovation with just a small variance. So that’s why I believe that Europol, for instance, here in Europe, is an essential hub for sharing.
And our collaboration with different countries, it’s great because we see the same needs in every country, perhaps with small differences, but still—and recently when we were in Brazil, same thing. I think this is universal and we need to, since we are, I think I agree with you, Oddvar, we are in this particular area, we are quite in the front.
And I think it’s our obligation to share this with our colleagues because what we do care about is helping people. That’s why we became police officers in the first place anyway. So this is our main task to help people. And if we can do that by sharing to help other colleagues doing the same thing, yeah, I cannot say it’s important enough.
BØRGE HANSEN: Very good.
BENTE SKATTØR: I would like to add also, because getting inspired by people which are much better than you, come on with it. Give us all, because that’s essential. And why shouldn’t we also be inspired by the criminals? Because they embrace the technology and they are really innovative. So of course we shouldn’t commit crime, but we should turn it around to combat crime.
Getting the inspiration of being, having the possibility to embrace—because I strongly believe if we embrace the technology, we can see the possibilities. We can break up and see the possibilities and believe me, we will anyhow handle within the rule of law. But this is instead of taking what is possible within the law, it’s a complete different thing.
BØRGE HANSEN: Do you think the criminals are eating elephants and running like that?
BENTE SKATTØR: They are running. And then they suddenly also get innovative in a strategic matter. Then it’s becoming organised crime. And they are really innovative also to get hold of money and also of course doing really bad things hurting people. But they are really innovative in that they actually—when they have a possibility, they’re actually going for the low-hanging fruits, and then they organise it very fast. So they’re doing business. And also, as many probably already know, in Norway, committing fraud using AI and technology is a bigger business compared to selling drugs in Norway now, at the time being.
BØRGE HANSEN: So let’s finish by looking forward and keeping it practical. If all this works, everyday investigative life and the quality of justice should improve. If you could change one thing in global policing tomorrow, one thing that would actually make investigators’ lives better and justice more reliable, what would it be and why? Cooperation?
BENTE SKATTØR: Cooperation across borders. Yes, together with academia, police, public sector and also industry across borders because the crime is borderless and we have to meet on the same, what you call, football stage as them. More cooperation.
BØRGE HANSEN: Cross-border cooperation.
KJELD HENRIK HELLAND-HANSEN: I agree quite much on that one. I mentioned the importance of sharing and exchanging ideas and seeing the possibilities to get that wider view to see what you can do. And also that you have leaders that see the importance of this, see, yeah, give you the space to try your wings, to try to be a bit more innovative, not just doing your day-to-day thing, but actually looking forward and see, have that vision. I think it’s important.
BØRGE HANSEN: It’s out of the proverbial box thinking. How about you, Oddvar? One thing you would want to improve.
ODDVAR MOLDESTAD: So learning more faster.
BØRGE HANSEN: So we’ve covered a lot, tech, practice, culture, law. But what I hear, it’s ultimately about people and how they work. Innovation in policing isn’t a future dream, it’s already here. And tonight, you’ve given us a glimpse of how it looks from the inside. Thank you for joining me.
ALL: Thank you.
END OF TRANSCRIPT
© 2025 Davidhorn. All rights reserved.


















