Why the threat is real
Ryan is a Threat Intelligence Specialist at Kodex. The keynote gives only his first name, for security. He spoke at the Kodex Global Keynote on October 27, 2026 about how Kodex keeps impostors off its network, and he started with what his team sees.
Stolen access is cheap. “I can go to a dark web forum right now and find police and government email logins for sale for as little as twenty bucks,” he said. A forged ID to go with the login costs an extra $40. He added that AI has threat actors inventing new and more sophisticated attack methods every day.
Six years of verification
For six years the team has had one job, keeping impostors off the Kodex Global network. In that time it has verified more than 160,000 investigators from over 15,000 agencies in more than 180 countries. Ryan said little about how. “If threat actors learn too much, they adapt, and the network gets harder to secure,” he said.
He did describe the newest part of the system. Citadel uses AI to detect threat actors among government users. This year the team brought in machine learning researchers and pointed them at Kodex’s data on confirmed threat actors, which covers thousands of them, plus dozens of technographic and behavioral signals tracked on top of their domains. Put in front of machine learning experts and experienced threat investigators, Ryan said, that data gives the detection that secures the network, potentially many times faster.
What a sign-up looks like
Ryan showed an investigator at a small regional force signing up at her station. “Citadel runs in the background while she fills in the form.” Her sign-up comes back clean and she is verified. Had anything looked wrong, it would have gone to a Kodex investigator, who would take it from there.
A verified account stays under watch. The investigator may need to be reverified later, for example to send emergency requests or to reach companies she wasn’t approved for.
“The moment something doesn’t add up, the account gets flagged and sent back for reverification.”
Ryan, Threat Intelligence Specialist, Kodex
Where it stands
Ryan said that by the end of 2026 every new sign-up will go through Citadel, so genuine investigators get in faster. Nobody sees or uses it directly, and the results show up in who gets verified.
For how Kodex checks the agency, the requester and the legal process, see How does Kodex verify law enforcement requesters? and the Kodex guide to verifying a law enforcement request. Ryan’s whole segment is in the Kodex Global Keynote.
In Ryan’s words
We have a reputation at Kodex for being the company’s most paranoid team. What we say to that is, just because we’re paranoid doesn’t mean they’re not out to get us. I can go to a dark web forum and find police and government email logins for sale for as little as twenty bucks. For an extra forty, you can get a forged ID to go along with the login. Thanks to AI, these threat actors are inventing new and more sophisticated attack methods every day. If you’ve seen what we’ve seen, you’d be paranoid too.
For six years, our team has had one job: keep imposters off the Kodex Global network. We’re a large international group of investigators, and we use a lot of tools and processes to keep threat actors out. We’ve verified more than a hundred and sixty thousand investigators from over fifteen thousand agencies in more than a hundred and eighty countries. I can’t say much about how we do it. If threat actors learn too much, they adapt, and the network gets harder to secure.
The system is evolving. You’ll never see it or use it directly, but you’ll see the results. We call it Citadel. It’s the world’s first application that uses AI to detect threat actors among government users. This year, we brought on a team of ML researchers and left them loose on what they love most, data. Kodex has the world’s largest set of data on threat actors using government accounts. We’ve tracked thousands of confirmed threat actors, and we track dozens of technographic and behavioral signals on top of their domains. Put that data in front of machine learning experts and experienced threat investigators, and you get the same best in class threat detection that secures the Kodex global network, but potentially many times faster.
This is an investigator at a small regional force signing up at her station. Citadel runs in the background while she fills in the form. Her sign up comes back clean, and she’s verified. If anything didn’t look right, her sign up would have come to us, and one of our investigators would take it from there. As long as her account stays in good standing, she won’t need to do this again. But that doesn’t mean we don’t keep tabs on her. She may need to get reverified later, to send emergency requests or to reach out to companies she wasn’t approved for. And Citadel keeps watching for signs of compromise. The moment something doesn’t add up, the account gets flagged and sent back for reverification. By the end of the year, every new sign up will go through Citadel. Real investigators will get in faster, and threat actors will still have to go looking for easier targets.
Edited for length and clarity. From the Kodex Global Keynote, October 27, 2026