New AI-native contenders are missing the real cybersecurity plot. What is obvious? > AI is forcing an architectural reset in security: attacks now have near-zero marginal cost and adapt constantly. > Adding an LLM is an incremental upgrade, not the reset the market actually needs. > Today’s security leaders (Palo Alto, CrowdStrike, even Splunk) were built to defend against human-scale, episodic attacks. > The next generation of primitives will be measured by speed of containment, not alerts or accuracy. So who can realistically build this and where do durable advantages exist when intelligence is cheap? Read more in @wquist and @yrechtman’s latest deck. Link in replies.