“The language models we have now are probably the most significant thing to happen in security since we got the Internet.”
Why Anthropic’s new model has cybersecurity experts rattled
The company says it has built its most dangerous model yet. Can its coalition of internet companies fix the internet before others catch up?
Nicholas Carlini - Black-hat LLMs | [un]prompted 2026
Anthropic's safety warnings may have just backfired — the government has pulled the plug on its most powerful AI | TechCrunch
Anthropic isn't hiding its frustration. "We disagree that the finding of a narrow potential jailbreak should be cause for recalling a commercial model deployed to hundreds of millions of people," the company wrote in a blog post.
A small number of samples can poison LLMs of any size
Anthropic research on data-poisoning attacks in large language models
Who’s Afraid of Chinese Models? (Stratechery Article 7-20-2026)
Claude Mythos #2: Cybersecurity and Project Glasswing
Anthropic is not going to release its new most capable model, Claude Mythos, to the public any time soon.
Jailbreaking Large Language Models: If You Torture the Model Long Enough, It Will Confess!
A Cautionary Tale…
Maintaining the love for coding in the time of AI
By Alex Band Over the last year, we have seen the profound effects Large Language Models (LLMs) have on our open-source software work at NLnet Labs. These can be grouped into two main areas: community contributions and security reports. Before we dive into this topic, let me first give
Anthropic's Mythos is a wake up call, but experts say the era of AI-driven hacking is already here | Fortune
Anthropic is limiting access to Mythos, but experts say similar capabilities are already in reach.
Security risks overshadow the debut of Europe’s X rival, W
W, Europe’s answer to X (formerly Twitter), has an unconventional approach to privacy, prompting security researchers to question how safe the platform really is.
Anthropic Drops Flagship Safety Pledge
In an abrupt shift, the company may release future AI models without ironclad safety guarantees
Inside the suddenly explosive world of AI safety
AI safety researchers warned that popular models would go rogue. This is only the beginning.
The web can be weaponised – and we can't count on big tech to stop it | Tim Berners-Lee
It’s dangerous having a handful of companies control how ideas and opinions are shared. A regulator may be needed
Anthropic on Twitter / X
Claude Fable 5 will be available again globally tomorrow.After a series of productive conversations with the US government, we're redeploying the model with a new set of classifiers to target and block more cybersecurity tasks. In the near term, some routine tasks like coding…— Anthropic (@AnthropicAI) July 1, 2026
The LLMentalist Effect: how chat-based Large Language Models rep…
The new era of tech seems to be built on superstitious behaviour
