For years, cybersecurity has been a game of cat-and-mouse played at a glacial pace. When a Business Email Compromise (BEC) attack hits, forensic investigators usually spend 7-10 days manually sifting through mountains of logs and email headers to find the "patient zero." It’s a tedious, error-prone process that can result in significant financial losses for companies in both labor and damages.

Asymmetric Security wants to flip that script. The San Francisco-based startup is emerging from stealth with pre-seed funding to prove that if AI is going to make attacks more sophisticated, it had better be used to make defenses faster.

The $55 billion problem

The timing for an "AI-native" forensics firm isn't just opportunistic; it’s arguably necessary. According to the FBI, BEC attacks have accounted for over $55 billion in losses over the last decade. A new era is beginning. As frontier AI models become more adept at mimicking human speech and social engineering at scale, the volume of these attacks is expected to rise.

"As AI progress accelerates toward AGI, the security stakes rise," said Alexis Carlier, co-founder of Asymmetric Security. "Autonomous AI attacks are no longer hypothetical. We exist to make AI an overwhelming advantage for defenders."

Asymmetric isn't just selling a software-as-a-service (SaaS) tool; they are positioning themselves as a "full-stack" AI services company. This means they provide both the proprietary AI platform and the expert human oversight required to handle high-stakes investigations.

How the "10x" efficiency works

The company claims it can shrink investigation timelines from days to hours. Here is how the platform handles a typical breach:

  • Autonomous Log Pulls: The system is designed to ingest evidence across email, identity providers, and cloud applications.

  • The AI Engine: It is designed to analyze large volumes of data to help reconstruct an attack timeline.

  • Human-in-the-loop: Expert responders (many of whom come from firms like CrowdStrike and Palo Alto Networks) validate the AI’s findings with the aim of supporting review in legal or corporate settings.

  • Proactive Defense: The platform produces actionable recommendations to harden the perimeter against the next attempt.

Asymmetric leverages the latest AI models to shrink these processes while maintaining the gold standard of accuracy.

Industry experience and strategy

The startup brings together professionals with backgrounds in technology, security, and incident response.

This deep academic and technical bench is part of a larger mission: Asymmetric doesn’t just want to react to attacks; they want to help frontier AI labs (like Anthropic or OpenAI) build better "evaluations" to prevent their models from being used for cyberattacks in the first place.

What’s next?

Asymmetric plans to use the fresh capital to aggressively hire engineers and incident responders. While the current focus is largely on BEC, the company is already eyeing expansion into more complex territories: insider threats, ransomware, and state-sponsored attacks.

As the "AI vs. AI" arms race heats up, Asymmetric is betting that the winner won't just be the one with the best model, but the one who can deploy it the fastest when the alarms start going off.

VentureBeat newsroom and editorial staff were not involved in the creation of this content.