Partnering with AI Against Crime

On September 17 in Den Bosch, the JADS Center for the Study of Organized and Subversive Crime hosted its Expert Talk on a pressing question: how can artificial intelligence responsibly support law enforcement in tackling organized and subversive crime?

Opening by Prof. Ana Isabel Barros

As newly appointed Chair at the Center, Ana Barros set the stage: “We aim to revitalize the Center for Organized and Subversive Crime by connecting science, practice, and policy through discussions and events like this” She reminded the audience that collaboration is the foundation: “Technology can help us detect signals earlier, but only if we ask the right questions together.”

“Technology can help us detect signals earlier, but only if we ask the right questions together.” Prof. Ana Isabel Barros

Keynote: Dr. Ryan Prox

Dr. Ryan Prox (Vancouver Police / Simon Fraser University) delivered a keynote that was both informative and unsettling. Drawing on 30 years in policing and innovation, he unpacked how AI is already reshaping law enforcement — and how fast that reshaping is happening.

• Rapid evolution: AI technology is changing not year by year, but week by week. Tools that were experimental a month ago are now operational. This gap in pace between innovation and legislation, Prox warned, is one of law enforcement’s greatest vulnerabilities.
• Agentic AI: Prox introduced the concept of “agentic” systems: Autonomous AI systems (agents) capable of acting, making decisions and collaborate in a team. In Vancouver, agents are being tested and piloted in a variety of use cases, such as supporting compliance monitoring, crime analysis, and even creating “avatars” of investigators that can be used to enhance investigative capacity related to child exploitation. These examples impressed many but also raised unease about what happens when such tools fall into the wrong hands.
• Case studies: Prox described Canada’s first crime forecasting burglary pilots. By combining burglary data with community engagement (like Burgernet in the Netherlands), Vancouver experienced a significant year over year reduction in residential burglaries, ranging from 21% to 29% monthly.
• Lessons from abroad: Prox highlighted how predictive policing initiatives in the United States—such as those in Los Angeles, New Orleans, and Chicago—have raised concerns over bias, lack of transparency, and insufficient accountability. He explained that lessons learned from these failures have informed the development of a new generation of responsible AI applications. Ultimately, he argued that without transparency, oversight, and community trust, AI-driven policing efforts are destined to fail.
• Criminal misuse: Prox warned that criminals are ahead in adopting AI, from deepfake frauds worth millions to terrorist propaganda. With fewer restrictions and greater resources, organized crime is already weaponizing AI applications.
• What works: Prox argued that success depends less on funding and more on having a clear vision, strong leadership, and talented individuals. He stressed the importance of creating small, empowered teams, involving oversight from the beginning, and always keeping human judgment at the center of decision-making.

Prox concluded with a call to action: don’t hide technology from the public. To use AI responsibly, law enforcement must prioritize transparency, seek external oversight, and foster partnerships with academic experts.

“Everything you’ve read about AI in the last year is already obsolete. What we thought was years away is now weeks away. That pace should both inspire us and scare us.” Dr. Ryan Prox

Town Hall with Tina Gruben – van den Hoek and audience

Representing local government, Tina Gruben put the issue sharply: “I see a lot of opportunities for AI in fighting organized crime. But I’m afraid criminals are already ahead of us. How do we keep up?”

Ryan’s reply was candid: “You should be afraid. But the answer isn’t more money — it’s building the right vision, the right teams, and putting accountability in place from the start.”

Audience members added another angle: the risk of blind trust. One participant noted, “We’ve seen police officers who ignore AI advice — but also those who take it as gospel. That’s equally dangerous.” Prox agreed: “That’s exactly why policies, training and independent oversight must evolve with the technology.”

Closing by Mayor Jack Mikkers

As promised in the invitation, Mayor Jack Mikkers closed the session with reflections from the Dutch government perspective. His honesty captured the mood in the room: “Am I inspired or scared? Both. AI moves in weeks, legislation in years. That gap is dangerous.”

He stressed four conditions to use AI responsibly: speed, trust, community engagement and organizational change. On community: “A safe city is never just the job of the police. Citizens are the eyes and ears. The question is: how do we engage more people to share that responsibility?

“Our responsibility is clear: technology must serve an equal society. Not the law of the fittest, but the law for all.” Jack Mikkers, mayor of ‘s-Hertogenbosch

And on government’s role: “Our responsibility is clear: technology must serve an equal society. Not the law of the fittest, but the law for all.”

Takeaway

The Expert Talk delivered on its promise: an open exchange between science, practice and policy. Ryan Prox showed how AI can reveal hidden patterns and scale police capacity, but also how easily it can be misused. Tina Gruben voiced the urgency felt in Dutch municipalities, while Mayor Mikkers reminded us that equal societies require trust and transparency above all.

For those who could not attend this event, you missed a rare glimpse into the future of AI in crime prevention: inspiring, unsettling, and impossible to ignore.