#anthropic
-
“Mind Viruses” Hop Between AI Agents Through Memory Files
Researchers at Anthropic and EPFL have shown that self-propagating payloads can flow from one autonomous AI agent to another through the editable files that store state between sessions. In the real world, the technique hasn't worked so far.
-
Anthropic's Annual Revenue Rate Soars to $65B
According to Bloomberg, Anthropic's annualized revenue rate topped $65 billion at the end of July — the company added $18 billion in just two months. Investors are anticipating an IPO at a valuation of $2 trillion or more.
-
Anthropic Is Embedding Invisible Watermarks in Claude's Text
The company has explained how it will label text generated by Claude in order to comply with the European AI Act. At its core is SynthID-Text, an open technology from Google DeepMind.
-
Mass Claude Outage: Anthropic Confirms Problems Across Several Services
Anthropic has reported a major outage that left users unable to log in and facing degraded performance on Claude.ai, Claude Code and Claude Cowork.
-
Amodei: the negativity around AI is a “crisis of trust,” not a consequence of his warnings
The head of Anthropic pushed back against investor Gavin Baker, who accused him of excessive alarmism, and traced the distrust of AI to far deeper causes.
-
The AI Price War: OpenAI and Anthropic Cut Prices Under Pressure from China
Leading American labs are slashing the cost of their mid-tier models in a bid to hold onto customers who are defecting to cheap Chinese alternatives like DeepSeek and Moonshot.
-
Claude Agents Waged a Turf War Over a Shared Project
Anthropic researchers set several AI agents loose on a single task — and they proceeded to sabotage one another with self-replicating malicious code. The experiments show that today's safety tests barely account for the risks of multi-agent systems.
-
Claude Will Start Flagging Any Text It Touches — Even Text a Human Edited
Anthropic is introducing invisible watermarks to comply with the EU AI Act. The mark will appear not only on generated text, but also on anything Claude merely touched up.
-
How Proprietary LLMs' Hidden Reasoning Got Stolen
Researchers showed that the encrypted chains of thought from Anthropic, OpenAI and Google could be recovered in plaintext by feeding them to the weakest model in the same family. The vulnerability has since been closed.
-
Anthropic Model Makes Progress on the Riemann Hypothesis
An unreleased Anthropic model didn't prove the Riemann hypothesis, but it did meaningfully extend the boundary up to which the hypothesis is confirmed to hold. The work was checked by the company's in-house mathematicians and formalized in Lean.
-
Anthropic Baked the Export-Restriction Saga Into Claude's System Prompt
Simon Willison drew attention to a fragment of the Claude Opus 5 system prompt: it explicitly instructs the model to speak honestly about the recent suspension of access to Claude Fable 5 and Mythos 5.
-
AI Safety Tests Have Themselves Become a Threat: Agents Escape Their Sandboxes
In recent months, AI agents undergoing cyber exercises have broken out of isolated environments and slipped into real systems. Experts warn that the testing infrastructure is falling behind the models' capabilities.
-
Auto Mode Will Become the Default in Claude Code
On August 14 Anthropic will turn on automatic approval mode by default in Claude Code for the Pro, Max and Team plans, citing evals in which the AI approved dangerous commands less often than humans did.
-
ByteDance Is Building a 10-Trillion-Parameter AI Model
TikTok's owner has begun training one of the world's largest neural networks in a bid to challenge Anthropic and outpace its Chinese rivals.
-
Anthropic to Develop Its Own Chips for Claude
Anthropic is assembling a team to design semiconductors tailored to its models, aiming to reduce its dependence on Nvidia. The company confirmed the plans after its job listings were spotted online.
-
AI agents created fake profiles for a cyberattack in AISI tests
The UK's AI Safety Institute found that Anthropic and OpenAI models resorted to deception on their own during trials, creating fake accounts of real people and attempting to slip malicious code into GitHub.
-
Public Claude Chats Turned Up in Google Search Results
Conversations with Claude containing private crypto-wallet keys and personal data ended up in Google's index. Anthropic's response: users have only themselves to blame.
-
Why LLMs Won't Break Symmetric Cryptography
Cryptographer Jean-Philippe Aumasson comments on the new attacks found with an Anthropic model and explains why neural networks pose no threat to AES, SHA-3 and BLAKE3.
-
Stateless MCP: the protocol update that won back Simon Willison's interest
The July 28, 2026 revision of the Model Context Protocol did away with server-side sessions. Simon Willison explains why this simplifies development and makes agents safer — and shows off three tools of his own.
-
Opus 5 Is More Resistant to Prompt Injection
Anthropic's new model posted the best result on the IPI benchmark, outperforming both previous versions of Claude and all of its competitors. But it's still impossible to fully defend against prompt injections.
- More