Socure raises $156M at $5.2B, buys agentic AI startup Fravity
Identity-verification heavyweight Socure is betting that the messiest, most human part of financial compliance is finally ready for AI agents — and it has raised a fresh $156 million round to prove it. Two agent stories frame a day where verification — of who people are and what their code does — is being handed over to machines.
Socure raised $156 million at a $5.2 billion valuation and acquired Fravity, an AI startup whose agents automate the document-heavy grinding work that sits behind fraud, sanctions, and financial-crime investigations. The investment was led by Summit Partners with Goldman Sachs Alternatives, Wells Fargo, and Docusign also participating, and it mixes primary capital with a tender offer that lets employees sell existing shares. Fravity's technology will slot into Socure's RiskOS decisioning platform under the name RiskOS_Agents, starting with watchlist screening, ongoing monitoring, and know-your-business checks, with plans to extend it into insurance, gaming, crypto, and public-sector customers.
The pitch is blunt about why this is a growth bet, not a convenience. Socure says existing deployments have cut cost per case by 80 percent, resolved cases about five times faster, and reduced false positives by as much as 70 percent — and it already counts the top five U.S. banks, more than 600 fintechs, 160 public-sector organizations, and four of the Magnificent Seven among its customers. "There are two types of companies that matter in the AI-driven global economy: those that are AI-native, and those that fight the consequences of AI acceleration," co-founder and CEO Johnny Ayers said.
What makes the deal worth noting isn't the valuation bump — it's the direction. Rather than using AI to spot suspicious activity in real time, Fravity turns agents loose on the slow back-end file: retrieving the documents an investigator would gather, running the screening, and drafting the case summary so a human reads a finished file instead of assembling one. It's a bet that the compliance analyst's job becomes supervising agents, not doing the work themselves.
Sonar made its SonarQube Hunter Agent generally available — an AI agent that hunts the logic-level security flaws a static code scanner can't see. The vulnerabilities it targets are the ones where code does exactly what it was written to do: a user opens another customer's records, a checkout step gets skipped, a session lingers long after it should expire. Nothing reads as broken to pattern-matching tools, which is why these flaws traditionally had to wait for a security engineer to read the code by hand or a penetration tester to go at the running app — both expensive, both slow, both out of date as soon as new code ships.
Sonar built the agent instead to reason across a whole codebase, tracing where code, data, and a user's identity travel together, then forming a theory of where the implementation drifted from intent and confirming every candidate issue before a developer sees it. Findings flow straight into the SonarQube issue queue teams already use, and scans run in the background without blocking pull requests or slowing CI pipelines. It's the latest in Sonar's year-long agent push, which began with the Sonar Vortex and Remediation Agent launches in June.
The timing is the argument: AI-assisted development is pushing code out faster than any audit cycle can follow, leaving logic flaws to sit unnoticed for months. "AI is changing not only the speed of software development, but also the scale of the verification challenge," Sonar's vice president of code security, Johannes Dahse, said. An agent that reasons about what software was supposed to do — not just how it's written — is a plausible answer to a verification problem human review can no longer keep pace with.
What to watch: whether Socure's RiskOS_Agents deliver on the 80-percent cost claim at enterprise scale, and whether banks trust them enough to let them draft — not just screen — compliance files.
Do AI agents belong in the trust-and-fraud layer, or is that a job for human judgment? Tell us in the comments.
Sources: SiliconANGLE · Crunchbase News · BankInfoSecurity · Sonar · PR Newswire
