AI versus AI: Five New Rankings in Verification and Security
Written by Wilco Verdoold
Every AI that can fake something creates demand for an AI that can check it. Five new rankings map the response: who you are, what you sign, who you hire, where you’re exposed, and what’s attacking you.
Generative AI made faking cheap. A passport photo that survives a first glance, a cloned voice on the phone, a phishing mail without the spelling mistakes, a CV with a work history that never happened. What used to take skill now takes a prompt.
The countermovement is just as real, and it is measurable. This month we add five new rankings to RankmyAI that together map a single chain of digital trust: who you are, what you sign, who you hire, where you’re exposed, and what’s attacking you. As always, these rankings measure popularity based on traffic, reviews, and investment. They show where the market puts its attention, not which tool performs best.
Who you are: ID verification & validation
The chain starts with the question every other check depends on: is this person who they say they are? MEGVII leads, ahead of Incode Technologies and Vectra AI. The detail that stands out: deepfake and liveness detection has become standard kit. Incode lists deepfake detection next to document checks, and AI or Not, which spots AI-generated images, audio, and video, ranks tenth. Proving a human identity now includes checking for machine-made fakes.
What you sign: e-signature solutions
Once identity is settled, agreements follow. The top of this list is not made of signature widgets but of contract platforms. Ironclad leads, with Brazil’s QI Tech and the UK’s Juro close behind, and for most of these tools the signature is the final step of a longer AI-assisted process of drafting, reviewing, and negotiating. Signing is turning into a feature rather than a product.
Who you hire: background checks
The smallest category of the five, and the most geographically distinct. Checkr and Fountain lead from the United States, but five of the top 25 come from Latin America, four of them from Brazil. Markets under high fraud pressure build their own verification tech, and this list shows it.
Where you’re exposed: vulnerability management (93 tools)
From people to systems. Tenable, Snyk, and Fing top a field of 93 tools that find weaknesses before attackers do. The newer story sits just below them: Promptfoo, a security testing platform for LLM applications, ranks sixth, and 0Din.ai, focused on generative AI vulnerabilities, sits in the top 20. The systems being scanned are, increasingly, AI systems.
What’s attacking you: threat detection
The last link is the active one: tools that watch attacks as they happen. Acronis leads, ahead of Cybereason and BigBear.ai, and the defense connection is no coincidence. BigBear.ai builds for defense, intelligence, and border security, part of the military AI landscape we mapped earlier. Deeper in the list a new layer appears: HiddenLayer, which protects AI models themselves, ranks ninth, with NeuralTrust in the top 20. Even the models now have bodyguards.
Five rankings, one pattern. The story is not that AI can fake a face, a signature, or a work history. It is that checking has become an industry of its own, and the current sizes show where trust is scarcest: 161 tools compete to prove who you are, against 55 for what you sign and 36 for who you hire. Every AI that can fake something creates demand for an AI that can check it. Both sides of that sentence are now in the rankings.
