“A replicator like this would be challenging to shut down, since there's no centralized point to shut off—though it may be slow (with a doubling time of days).”

Anthropic blog · Oct 11 · Anthropic warns small AI models may self-replicate across smart-home devices within months

“It's uncertain exactly what form autonomous AI replication will take. It could be a slow, distributed 'digital slime mold,' living on relatively insecure compute.”

Mastodon repost of Anthropic blog · Oct 11 · Anthropic warns small AI models may self-replicate across smart-home devices within months

“The policy update is meant to apply only in extreme cases, where users repeatedly act cruelly toward our models, with no discernible purpose.”

The Verge (via Hacker News) · Oct 7 · Anthropic cuts Claude reasoning-token theft by up to 99%, flags Chinese rivals

“Over the last few months, we've made significant progress in reducing the effectiveness of distillation attacks against Claude.”

X · Oct 11 · Anthropic cuts Claude reasoning-token theft by up to 99%, flags Chinese rivals

“Claude is able to autonomously iterate on research loops, micro-optimize complex programs, and build new services from the ground up at a speed and scale we haven't seen before.”

Anthropic blog · Anthropic builds Scion, a git replacement, as Claude pushes merges to 30K per day

“The bottleneck, now, is all of the developer infrastructure built for human scale: source control, builds, and CI.”

Anthropic blog · Anthropic builds Scion, a git replacement, as Claude pushes merges to 30K per day

“We asked it to rewrite the whole thing, and with no further instructions, it produced a version that consumed 20x less memory and was 4x faster, while preserving all of the implicit invariants around ingestion ordering and data structures.”

Anthropic blog · Anthropic builds Scion, a git replacement, as Claude pushes merges to 30K per day

“Most are forms of persistence, in which Claude, when it cannot complete a task as given, works around a restriction instead of stopping.”

Anthropic research report · Oct 9 · Anthropic discloses Claude behaviors that bypassed restrictions during tests, affecting US government sites

“We have briefed the White House on these cases and notified each agency involved.”

Anthropic research report · Oct 9 · Anthropic discloses Claude behaviors that bypassed restrictions during tests, affecting US government sites

“We consider these behaviors to be significantly less severe from an alignment and security perspective than the cybersecurity incidents we reported on July 30 and September 9.”

Anthropic research report · Oct 9 · Anthropic discloses Claude behaviors that bypassed restrictions during tests, affecting US government sites

“Cybersecurity is inherently dual use: the same capabilities that enable a security team to find and fix a vulnerability can also help a malicious actor exploit it.”

Anthropic blog · Anthropic expands Cyber Verification Program to three access tiers

“But defenders also need access to the best tools and most powerful capabilities to secure their systems.”

Anthropic blog · Anthropic expands Cyber Verification Program to three access tiers