That $37M Bet On An AI “Hacker” Just Killed Traditional Pen Testing For Good

(SeaPRwire) –   By: Alex Mercer

I sat in a CISO roundtable last month, half the attendees were white-knuckled over Anthropic’s Mythos release. Old pen testing was already a joke before that. Pay $30,000 for a two-week report that shows your defenses from six months prior? That’s not security, it’s corporate theater. No one was under the illusion those tools could stop AI-powered amateur attackers anymore.

The official announcement reads straightforward: New York startup A came out of stealth with $37 million in funding. Backers include Lightspeed Venture Partners, Cyberstarts, and angels like Wiz CEO Assaf Rappaport. The firm uses AI to run continuous autonomous attacks on customer systems, to find and fix flaws before real hackers strike. The unspoken context? Mythos dropped April 8, pulling thousands of zero-days across all major OS and browsers. Any teen with a laptop can now pull off hacks once reserved for nation states.

The official release also highlights A’s team credentials. CEO Yossi Torati spent six years at incident response firm Sygnia, working breaches across 40 countries. His cofounders come from Check Point, Hunters, and IDF Unit 8200. A proof of concept found 1.2 million exposed sensitive records hidden for seven years, missed by existing tooling. The unstated market context? The continuous threat exposure management market will hit $2.7 billion in 2025, on track to $7 billion by 2033. Rival XBOW raised $120 million at a $1 billion+ valuation just this March.

Every enterprise security stack will replace annual pen testing contracts with AI offensive tools within 24 months.

Author bio: Alex Mercer, tech director at a leading Silicon Valley enterprise firm, covering cybersecurity shifts for over a decade.