A new guide showcases how to build a secure, portable AI coding agent using Docker Compose, MCP Gateway, and Cerebras Cloud. It also demonstrates how a no-network code sandbox can be created using Testcontainers to protect developers’ workflows.

Anthropic’s new “Claude for Chrome” tool marks a limited but notable step toward browser-based AI integration. Its debut comes at a time when security concerns surrounding AI-powered browsers are front and center.

Google Cloud has unveiled an “agentic” Security Operations Center meant to accelerate incident management while rolling out new protections across its product suite. The move, announced on August 19, broadens the company’s strategy for securing artificial intelligence and reinforcing trust in its cloud services.