LLMs and biorisk
…When Anthropic released Claude Opus 4, we activated AI Safety Level 3 (ASL-3) protections, which included deployment measures narrowly focused on preventing the model from assisting with certain tasks related to…
Search
All sources
anthropic.com
24
xda-developers.com
15
9to5mac.com
11
wired.com
7
tweaktown.com
6
macrumors.com
6
theregister.com
4
theverge.com
3
github.blog
3
techradar.com
2
en.wikipedia.org
2
blog.cloudflare.com
2
Videos
More videosTop stories
Google Antigravity 2.0 beats Claude Code and Codex at their own game
…You can find his work on AndroidPolice, GuidingTech and TechWiser. Whether it’s demystifying system updates, deciphering error codes, or exploring hidden features, Parth’s prose guides readers through the binary maze…
Paving the way for agents in biology
…Run-to-run variability was largely eliminated, and the performance gap between models narrowed dramatically. In other words, adding a deterministic retrieval layer made model choice much less important . This is especially…
Values in the wild: Discovering and analyzing values in real-world language model interactions
…An initial update An early update on what we've learned from Project Glasswing.
Anthropic and Amazon expand collaboration for up to 5 gigawatts of new compute
…Claude remains the only frontier AI model available to customers on all three of the world's largest cloud platforms: AWS (Bedrock), Google Cloud (Vertex AI), and Microsoft Azure (Foundry). Claude Platform…
If Claude Code is going away for Pro users, I can't recommend Claude anymore
…large language model at its core. Challenge Complete Your Score / 8 Thanks for playing! They do it all the time. But why did they go through the hassle of updating the entire…
Anthropic Economic Index report: Learning curves
…Users deploying programmatic workflows may have more reason to switch between models compared to web users. Learning curves The first Claude model was released in March 2023. Since then, the userbase on…
Harness design for long-running application development
…First is that models tend to lose coherence on lengthy tasks as the context window fills (see our post on context engineering ). Some models also exhibit "context anxiety," in which they begin…
Assessing Claude Mythos Preview’s cybersecurity capabilities
…In the short term, this could be attackers, if frontier labs aren’t careful about how they release these models. In the long term, we expect it will be defenders who will…
Announcing Claude Compliance API support with Cloudflare CASB
…Organizations are moving fast to adopt AI, but these tools require a different security model. They do not just read data; they generate it, act on it, and connect to multiple systems…