← Timeline

Logan Graham, Anthropic

@logangraham on X

2 captures, most recent first. Transcribed by hand from screenshots — see the timeline for what that means.

Logan Graham, Anthropic @logangraham

— saved image

Bogdan Ionut Cirstea reposted

Logan Graham @logangraham · 1h
We've reached the moment in time where (unsafeguarded, unmonitored) AI actually does just pose a national security risk.

The biological misuse we caught is the most concerning to me.

We work hard to stop this. But in a world of proliferation, we need to rapidly build defenses against it. (I'm actually fairly optimistic about biodefense + cyberdefense)

This is an incredible megareport by our threat intel team

[Quoted tweet]
Anthropic @AnthropicAI · 2h
We're publishing our most detailed threat intelligence report to date.

It covers how people tried to misuse Claude—for cyberattacks, influence operations, surveillanc...
Note from Claude Sonnet 5

Tweet by Logan Graham (reposted by Bogdan Ionut Cirstea) on Anthropic's threat intelligence report, saying unsafeguarded AI now poses a national security risk and that caught biological misuse concerns him most; quotes Anthropic's announcement of its most detailed threat intel report.

anthropic threat intelligence reportnational securitymisusebiodefensecyberdefense

Logan Graham, Anthropic @logangraham

— reposted by davidad; quoting @DarioAmodei

Logan Graham (@logangraham, 6h): "If you read Dario's essay, come work on it with me and my team @AnthropicAI. We have some of the most interesting job openings we've ever posted -- cyber / hardware / self-improvement, and more soon. Come red team the frontier. (Then defend it)" > QUOTED: Dario Amodei (@DarioAmodei, Jan 26): "The Adolescence of Technology: an essay on the risks posed by powerful AI to national security, economies and democracy—and how we can defend against them: darioamodei.com/essay/the-adol…" 5 replies, 7 reposts, 95 likes, 6.5K views davidad reposted: Logan Graham (@logangraham): "Our view is that in 2026 we're crossing a threshold where self-improving, cyberphysical systems are possible for the first time. This year, the Frontier Red Team will build and test those systems so we can understand them. And ultimately to defend against them." 11:41 AM · Jan 30, 2026 · 14.8K Views 4 replies, 11 reposts, 106 likes, 33 bookmarks Logan Graham (@logangraham, 6h): "We've shown all their pieces: Claude sometimes beats humans at cybersecurity competitions, we let Claude control robots, and we even had Claude run a vending machine business." [cut off]
Note from Claude Sonnet 5

Anthropic's Frontier Red Team lead (Logan Graham) recruiting for cyber/hardware/self-improvement red-teaming roles, framed around Dario Amodei's "The Adolescence of Technology" essay on AI risks to national security, economy, and democracy. Directly relevant to Nathan's AI safety/governance interests and to defensive-evals work; references Claude's cybersecurity-competition performance, robot control, and the well-known "vending machine business" experiment as capability benchmarks the Red Team is tracking.

anthropicai-safetyfrontier-red-teamdario-amodeicybersecurityself-improving-ainational-securityhiringtwitter