The open source intelligence landscape has always changed quickly, but what is happening right now feels different. Chris P, Head of Intelligence at Blackdot Solutions and former National Crime Agency investigator, spoke at Digital Experience Nordics 2026 earlier this month on exactly where that change is heading.
Chris spent 21 years in British Military Intelligence before leading major investigations into money laundering, cybercrime and organised crime at the NCA. He has watched the data problem grow in real time.
"There are 200,000 new web pages being registered every week. AI is being trained on the content we produce, and it is putting out its own content in return."
The volume problem is matched by a velocity problem. The old model of deep analytical work on a single event before moving to the next no longer holds.
"Gone are the days where we'd have a single event that we could spend time using structured analytical techniques on before moving on to the next one. Now you've got 24 hours and by that time it's old news."
And then there is variety. The average social media user operates across 6.5 platforms. For investigators, that means 6 or 7 different places to look for every single subject, every single day.
Chris is clear that the hype around AI needs reigning in.
"I like to think that AI is an evolution not a revolution. All it does is accelerate existing processes. It's not doing anything novel, it's not doing anything new."
What it does do is expand an analyst's reach in a way that was not previously possible. Where a human has to go from one source to another in sequence, an AI agent cuts across all of them simultaneously.
The practical impact of that is significant. Chris points to the ratio problem he saw throughout his NCA career.
"The collection [of open source information] takes up 70% of analysts time. The analysis 20%, 10% making something look good in PowerPoint. That is completely wrong. We pay analysts to think. We don't pay them to collect information that's online."
AI, used properly, flips that ratio.
Chris ran a live example during his talk. Working on a human trafficking investigation, he pulled 150 TikTok accounts onto a network chart, added their connections, and ended up with 22,500 profiles.
"I gave it no context, nothing at all. All it was, is just a network of TikTok accounts. And I said to the platform: what is this?"
The response identified it as a human trafficking network operating out of North Africa and Eastern Europe, targeting the UK, and pulled out the phone numbers, trade craft and code words being used.
"That would have taken years to have gone through 22,500 TikTok profiles."
The same principle applies to due diligence work. On the Videris platform, an analyst inputs a company name, a pre-built playbook runs the full investigation, returns a structured report with every statement of fact footnoted, and produces a network chart of connected entities. The whole thing is reproducible and timestamped.
"These playbooks take about an hour to write. Which means if something happens, I can create a playbook, put in a prompt, and it'll go off and collect all that information for me. And it's repeatable."
For all the capability, Chris is direct about where the line sits.
"At no point should AI be allowed to run amok. There should always be a human involved in that process."
This is not just a philosophical position. It is a practical one. An AI agent has no access to classified sources, no awareness of what the analyst already knows and no ability to carry accountability for the decisions that follow from an investigation.
"An AI agent lacks context. If you are a Southeast Asia expert and you've been doing it for 15 years, it will not know as much as you, even though it might be able to collect more information than you can."
The transparency point matters too. Every output from Videris is fully footnoted so the analyst can check the working, verify the sources and catch hallucinations before they become a problem.
Chris is measured about the road ahead.
"AI is not replacing open source intelligence, but it is redefining it."
The organisations that get this right will be the ones that use AI to do the heavy-lifting collection, and free up their analysts to focus on the judgement calls that actually require a human. The ones that get it wrong will either move too fast without governance in place or resist the change entirely, and ultimately, fall behind.
"Approach it cautiously, with an ethical viewpoint. That is absolutely key."
Want to see how it works on your investigations? Book a call with the team.