Artemis raises $70M to build AI-native security platform that fights machine-speed attacks
What's the deal? Artemis, an AI-native cybersecurity startup, has emerged from stealth with $70 million in combined seed and Series A funding, just six months after founding. The Series A was led by Felicis, with First Round Capital and BrightmindDealroom has a profile for this one. Try Dealroom → returning alongside Theory VCDealroom has a profile for this one. Try Dealroom →, Two Sigma, and LockstepDealroom has a profile for this one. Try Dealroom →. Angel backers include the founders of Demisto and Abnormal AI, the former chief executive and chief technology officer of Splunk, and senior executives from CrowdStrike, Palo Alto Networks, Microsoft, and Okta.
The company is already processing billions of events per hour for enterprise customers in technology, banking, and financial services.
Why now? AI-powered attacks now execute in minutes, adapt in real time, and never repeat the same pattern — rendering rule-based detection, which takes a week to write and ships outdated, structurally inadequate. Artemis was built to match that speed: its platform generates customer-specific detections autonomously, investigates every signal, and surfaces correlated attack narratives rather than disconnected alerts. One early customer reduced investigation time by 96%, completing cases in under five minutes.
The founding team brings directly relevant pedigree. Chief executive Shachar HirshbergDealroom has a profile for this one. Try Dealroom → built Demisto — which created the security orchestration, automation, and response (SOAR) category before being acquired — and led GuardDuty at AWSDealroom has a profile for this one. Try Dealroom →. Chief technology officer Dan ShieblerDealroom has a profile for this one. Try Dealroom → led machine learning and AI at Abnormal AI after a PhD in machine learning from Oxford.
What could go wrong? Artemis is entering one of the most competitive markets in enterprise software, against incumbents including CrowdStrike, Palo Alto Networks, and Microsoft — all of which are also building AI-native detection capabilities. The company is six months old with no public track record at scale. Its data model approach, which retrieves data on demand rather than ingesting it upfront, is a meaningful architectural bet — but one that needs to prove reliability in production across diverse enterprise environments.
The signal: Artemis's $70 million raise at seed and Series A stage, six months in, reflects the urgency investors are attaching to the AI versus AI security problem. As AI lowers the cost and complexity of launching sophisticated attacks, the defenders who can match that speed — automatically, at machine scale — will define the next generation of enterprise security. The company's claim to deliver value on day one, driven by word-of-mouth demand before any marketing, is an unusually strong early signal in a category where enterprise sales cycles are notoriously long.
A.M.
