The future of identity: who’s really on the other end?

At this year’s Veriff Summer Week flagship event, Grand ‘Hey, Veriff’, moderator Tom Pooley, Data Product Manager at Veriff, sat down with Kaarel Kotkas, Founder and CEO, Hubert Behaghel, CTO, and Rob Brazier, CPO, for a fireside chat on The Future of Identity – digital IDs, the AI-fraud arms race, and what happens to identity when the “user” on the other end of a transaction might be an agent instead of a person.

Digital identity’s decade is finally arriving

Asked which trend was top of mind, Rob Brazier didn’t hesitate: digital identity. It’s a space that’s been talked up for years without matching action, but the pieces are now genuinely coming together. “It’s one of those trends that has been hyped for so many years that we’ve all grown quite bored of it, but that’s normally when it actually starts to take hold and matter,” he said, predicting digital identity goes mainstream for most of the world within the next couple of years.

Kaarel Kotkas framed the opportunity in practical terms by bringing everyday examples: using Klarna from the sofa without a wallet in reach, or proving you’re really you on a call with an Uber driver rather than an agent impersonating one.

For Veriff, he said, the value isn’t just verifying an identity once – it’s building a trusted identity that carries biometric authentication and risk signals with it, so it can be reused as a source of truth across a marketplace of interactions.

Kaarel Kotkas, Founder & CEO of Veriff

Fraud, AI, and a wake-up call

Hubert Behaghel kept his focus on fraud, and specifically on how AI is arming both sides of the fight. “Fraud as a service” now lets attackers find new attack vectors faster and exploit them at unprecedented scale, for example, injection attacks. His answer is Veriff’s multi-layered approach: document checks, computer vision, device integrity, and network intelligence working together, so beating one layer doesn’t mean beating the system. He also pointed to the speed advantage of owning the full stack: “What takes months for an IDV provider who buys a solution from a vendor – takes days for us.”

Kaarel brought the stakes home with a story from a visit to his home island, watching his own father spend half of their limited time together on the phone with someone attempting an insurance-related scam. It’s a reminder, he said, of the gap between how fast fraudsters iterate and how slowly the market, from banks to individual users, catches up on education and tooling. Closing that gap is what keeps him up at night.

Digital IDs won’t kill fraud, they’ll relocate it

Turning to what digital IDs mean for fraud, Rob’s view was that they’ll push fraud toward the edges: harder to forge a digital ID, but because rollout is so uneven globally, physical documents and the fraud that targets them aren’t going anywhere soon. “Fraud is a data and signals game,” he said.

Hubert added a wrinkle: a “pre-verified” digital ID is really just an assertion that verification happened at some point in the past, not proof that the person presenting it now is the same person. “Let’s be honest, there is no more just verification; there is an assertion that it’s been verified in the past. But now you have a problem, because you’re trusting the device, and you can’t.” That’s why he sees liveness checks becoming second nature, not a one-time gate, but a recurring check that the human behind the device is still the accountable human.

Hubert Behaghel, Chief Technology Officer at Veriff

Verifying agents: real problem, not yet a real product

The panel dug into one of the thornier questions in the room: what happens to identity when a bot, not a person, is doing the transacting? Kaarel noted that despite the buzz around agentic commerce, there is more agentic-than-agentic commerce today: “Can I trust this agent? Can this agent make transactions at a set level? That will be a key topic.”

Rob added that before businesses can verify a bot is acting on a consumer’s behalf, consumers need to be delegating meaningful purchasing decisions to agents at scale, and that shift hasn’t happened yet. If it follows the pattern of past internet shifts, he said, agentic commerce will likely aggregate around one or two dominant players who end up dictating how identity and consent get passed to an agent. “I think there’s a bunch of technology problems to solve. Once those get sorted out, the product problems will emerge, but they haven’t yet.”

What’s next in verification

Asked to pick a favorite engineering bet, Hubert pointed to passive verification: building enough signal (email, phone number, device history, IP history) to establish trust without ever interrupting the user. Some of the more advanced platforms on the internet today, he noted, are already moving in this direction, using passive signals to reduce anonymity without adding friction to the user experience. He also flagged the growing importance of orchestration: as Veriff bundles documents, digital wallets, and self-asserted IDs into a single flow, accuracy has to extend beyond the check itself to choosing the right method for the right user, automatically.

Rob was less willing to name a single favorite – “it’s like asking a dad to pick his favorite child” – describing a roadmap that balances near-term priorities with longer-term bets, including a deeper move into KYB (Know Your Business) and continued investment in passive verification over time.

Rob Brazier, Chief Product Officer at Veriff

The bigger picture: trust is the new barrier to entry

Closing the conversation, Kaarel reflected on ten years of building Veriff and what still surprises him: not any single technology shift, but how much bigger the opportunity keeps getting. With tens of millions of businesses and billions of people online, and verification needs recurring across services every year, he argued the industry is still only scratching the surface. “The barrier of entry is not payments anymore,” he said. “It’s a barrier of trust. And if we can be the one that solves it, it’s one of the biggest opportunities in technology.”

Glass half full, as Kaarel put it – bring on the next twenty years.