What it takes to get listed in GOFINDME: the editorial criteria, explained

When GOFINDME launched in May 2026 with 27 listings across 8 categories, the number felt deliberately small. A directory of European AI business tools had just gone live, and it contained fewer entries than most people's bookmarks folder. Well, that was precisely our point.

The EU's market for AI-powered business software is not short of products. Vendors do their best to place themselves infant of buyers with good descriptions of what they do and who they are, Many landing pages claim compliance with regulations that their legal teams are still meeting about. A tool built by a team in a European city, say, Hamburg, gets categorized the same way as one assembled overnight on top of a U.S. API wrapper with a EU domain registered few days ago. For a buyer trying to make a procurement decision, all these distinctions matter enormously and are almost impossible to verify from the outside. Which is why we believe what is really missing is clear signal about the AI powered infrastructure growing across the continent.

GOFINDME is Prudos's answer to that problem, and the rigorous editorial criteria behind it are what make it one thing rather than another.

The foundation of those criteria is a principle that sounds simple and turns out to be demanding: a product placement in GOFINDME is not for sale. Prudos publishes this as policy. A company cannot buy its way into the directory, cannot pay to move up in results, and cannot purchase a badge. This is not standard practice in the directory and software review space. G2, Capterra, and their peers operate on models where vendors pay for premium placement, enhanced profiles, and category visibility. A business model that has a critical catch: while the resulting lists are not necessarily dishonest, they are shaped by marketing budgets as much as by product merit. GOFINDME operates on a different premise, one that requires the editorial team to absorb the cost of curation rather than offload it to the companies being evaluated.

Curation that begins with rigorous verification of the fields that most directories take on faith. When a vendor claims EU hosting, Prudos checks it. When a tool claims GDPR compliance, the editorial review asks for the basis of that claim, because "we're GDPR compliant" on a website means almost nothing without knowing whether data processing agreements are in order, where subprocessors are located, and whether a Data Protection Officer has been designated where required. The EU AI Act has added another layer to this. Tools that fall under its scope are assessed for their claimed risk tier, their conformity obligations, and whether their documentation is consistent with what the Act actually requires. This is regulatory reading work, not marketing reading work, and it tends to produce a different picture than what the market has been used to.

The EU AI Act Readiness badge is the most visible output of this process. It is awarded after editorial review, and it signals that a tool's claims about its regulatory posture have been examined rather than accepted at face value. Getting it requires a company to engage with the review seriously. For buyers, it functions as a proxy for the due diligence they would otherwise have to conduct themselves, which for most procurement teams at small and mid-sized European businesses means due diligence that simply does not happen.

The Founding Member program gives some shape to how the directory has grown from that initial 27. The first cohort is capped at 50 companies, with the first year free and a fee of €499 per year from year two. We chose this structure to give businesses with strong products but limited marketing budgets the same access to editorial review as better-funded competitors. A bootstrapped analytics company like Plausible Analytics, which is self-funded, open-source, and EU-hosted, competes for attention in the same space as enterprise software that has raised nine-figure rounds. Cognigy raised $100 million in its Series C from DTCP in 2022. n8n closed a €156 million Series C with Insight Partners in 2024. The directory's job is to make the funding level irrelevant to the editorial outcome, so that a company with a genuinely good product, an energized team and a genuine EU compliance posture gets seen, regardless of whether it has a VP of Marketing running a review-generation campaign.

What gets a company listed, then, is a combination of things. The product has to be real in the sense that it is being actively developed and used. The company has to be able to substantiate its claims about where data lives, how it is processed, and what its obligations are under applicable EU regulation. The editorial team has to be able to reach a confident view on the tool's category, its primary use case, and whether the claims made about it are consistent with what the product actually does. Companies that describe themselves in the most expansive possible terms, with AI featured prominently in contexts where a closer look reveals a thin integration layer, do not pass this last test well.

There is something worth naming directly here. Verifying regulatory claims about EU AI tools is genuinely hard. The EU AI Act is a long and technically demanding text, and the guidance around it is still developing. Applying it to a specific product requires judgment calls that even specialists disagree about. Prudos does not claim to resolve every single ambiguity or to function as a legal entity. What the editorial criteria commit to is something more nuanced and more honest: reading the source material, asking the questions that the source material raises, and publishing the results of that review rather than the results of a vendor's marketing copy.

That commitment is what distinguishes the GOFINDME editorial criteria from the alternative, which is a list that gets longer because getting longer is good for the business that runs it. Personio, valued at €8.5 billion in its 2022 Series E, has the resources to appear prominently in any directory it chooses to invest in. The point of an editorially curated list is that prominence and investment do not determine the result. The editorial record does.

For buyers, this means a directory where a smaller number of verified listings is more useful than a larger number of unverified ones. For companies seeking inclusion, it means the preparation required is the same preparation a serious buyer would conduct during procurement. We designed the curation criteria for the GOFINDME listings to align with that.