Artemis emerges from stealth with $70M to build an AI “brain” for cybersecurity

Big raise, big ambitions

It has been reported that Artemis — a six-month-old startup that wants to replace rule-based cybersecurity stacks with a centralized, AI-driven “brain” — has closed a $70 million Series A led by Felicis, with participation from returning backers First Round Capital and Brightmind, plus Theory VC and a who’s-who of security veterans. The company’s founders, CEO Shachar Hirshberg (ex-AWS product lead) and CTO Dan Shiebler (formerly head of AI at Abnormal Security), say the product already has attracted customers. It has been reported that clients include Mercury, Wix, Lemonade and Abnormal AI, and that Artemis “has closed a few seven-figure deals” and expects multi-million dollar ARR before the end of 2026.

A single system to spot—and stop—AI-speed attacks

Artemis’ pitch is simple and urgent: attackers are using AI to move at machine speed, and defenders glued to brittle, rule-based tools are getting steamrolled. It has been reported that the product continuously monitors logins, cloud activity, apps and more to learn an organization’s “normal,” then connects the dots into a coherent incident narrative and can automatically take action — for example, locking a compromised account — before an intrusion spreads. “This isn’t just about the future getting worse,” Hirshberg told reporters. “The capabilities that exist today are already incredibly powerful, and attackers are leveraging them right now.” Scary thought, right? Especially when a single misstep can cascade in minutes.

Positioning, rivalry and the road ahead

Investors argue Artemis is riding a larger cycle: after years of tool sprawl, AI is nudging security back toward consolidation — a brain that ingests data, reasons across silos, and acts in real time. Felicis partner Jake Storm even framed Artemis as a next-gen alternative to legacy platforms like Splunk (acquired by Cisco in 2024 for $28 billion). But the field is crowded and fast-moving; nearly every major vendor is racing to bolt AI into detection and response, and startups promise autonomous defenses with increasing frequency. Anthropic’s Mythos preview and CrowdStrike reports showing shrinking “time to attack” only add urgency.

Will centralization work where patchwork failed?

Artemis is betting that automation plus contextual reasoning can un-jam security operations and give defenders back precious seconds. The wager makes sense on paper. Execution will be everything. Can one platform really replace a decade of stitched-together tooling and human workflows? That’s the question investors are betting on — and attackers are counting on the answer being slow. Time will tell.

Sources: fortune.com