Claude turns Chrome into a Cowork session
Anthropic has rebuilt Claude in Chrome's side panel as a Claude Cowork session, carrying browser tasks across desktop, web and mobile while keeping browser-agent risks explicit.
Security News and Analysis
6 verified stories covering Security, product updates and industry developments.
Coram raises 35 million dollars for AI physical security
Coram is applying autonomous video investigation to cameras, access logs and visitor data, moving security systems from recording evidence to answering questions.
OpenAI hardens Codex with a Windows sandbox
The new design uses separate Windows accounts, network policy and a command runner to make Codex isolation an operating-system boundary rather than an environment-variable promise.
OpenClaw open-source framework faces supply chain security trust crisis
Malicious typosquatting packages on npm impersonated OpenClaw's official dependencies, exposing supply chain risks in open-source AI frameworks and prompting the team to strengthen security measures.
GPT-5.3 Codex Receives OpenAI's First High-Severity Security Rating
OpenAI's GPT-5.3 Codex model has been assigned the company's first "high" cybersecurity risk level under its internal safety assessment framework, prompting the deployment of the most comprehensive security protections to date.
Opus 4.5 Introduces New Approach to Prompt Injection Defense
Anthropic's Opus 4.5 model improves prompt injection resistance through system prompt prioritization, context isolation, and refined refusal strategies, though no model is fully immune.