The internet has a trust problem. It's not new, exactly, but generative AI has turned a slow leak into a flood. When anyone can produce plausible-sounding text, images, and video at scale, the gap between what's published and what's true widens by the hour. Copenhagen-based Flare thinks infrastructure, not policies, is the answer.

Flare closed a EUR 3.6 million pre-seed round led by 20VC and 20Growth, the fund and growth firm run by Harry Stebbings. byFounders co-invested alongside a syndicate of angels from Stack Overflow, GitHub, Reddit, Meta, Kahoot!, HubSpot, and Encord.

That angel list is worth pausing on. Stack Overflow, GitHub, Reddit. These are the platforms that built the internet's existing knowledge infrastructure. Their people are investing in what they think comes next.

Flare isn't a fact-checking service. CEO Nicolai Frost Kolborg Jacobsen is building something more structural: a knowledge validation layer that evaluates claims at scale through transparent, collective intelligence. Think of it as the plumbing beneath a healthier information ecosystem, not the editorial judgment on top of it.

"What's broken right now isn't that people are naive," Jacobsen said. "It's that the infrastructure for discernment hasn't scaled with the infrastructure for content. That's what we're building."

The distinction matters. Content moderation is a human-intensive, politically fraught, fundamentally unscalable approach to information quality. Flare's bet is on transparent mechanisms that indicate whether a claim is supported, disputed, or unverified, without requiring a central authority to make the call.

Flare's founding team brings experience from Google, Unity, IBM, Amazon, and Corti, the Danish AI company building real-time clinical decision support. They've worked on systems where scale, quality, and trust had to coexist, which is a surprisingly rare combination of experience.

The pre-seed funding will go toward platform development, team expansion, and bringing the trust infrastructure to market. At EUR 3.6 million, it's early-stage money for what Flare frames as a civilizational-scale problem. The ambition-to-capital ratio is enormous, which either means the team is delusional or the approach is so fundamental that a small team can make disproportionate progress.

The timing isn't accidental. The EU AI Act's transparency requirements for AI-generated content are creating regulatory demand for exactly the kind of provenance and verification infrastructure Flare is building. European companies that solve this problem have a natural advantage: proximity to regulators, cultural comfort with trust frameworks, and a market that's mandating solutions rather than just hoping for them.

Copenhagen, specifically, has been developing a cluster of companies working on the intersection of AI and trust. Corti in healthcare. Pentia in digital ethics. Flare in information verification. It's not a coincidence. There's something in the Danish approach to technology, pragmatic, institution-minded, slightly skeptical of hype, that produces companies focused on trust rather than disruption for its own sake.

Building trust infrastructure for the entire internet sounds like the kind of project that needs a billion dollars and a decade. Flare is starting with EUR 3.6 million and a hypothesis. The gap between those two things is where startups either find product-market fit or discover that the problem was bigger than they thought.

But the alternative, an AI-saturated internet with no scalable mechanism for distinguishing signal from noise, isn't really an option. Someone has to build this. The people who built Stack Overflow, GitHub, and Reddit think Flare might be the ones to do it.

Given what those platforms know about information at scale, that's not nothing.