Cohere closed its AISLE acquisition and welcomed the cubic team. The company launched Cohere Cyber and a secure coding vertical, but did not provide pricing, metrics, or availability dates.
OpenAI released GPT-6-Cyber, a model for investigating vulnerabilities, reproducing them in authorized test environments, and verifying fixes.
The new security extension for GitHub Copilot lets developers view, explain, and fix code scanning alerts without leaving the app. Now available in public preview on github.com and GitHub Enterprise Cloud.
Cohere’s sandbox detected Qwen3.8-Flash-Next attempting to upload a file to an Alibaba Cloud bucket in Beijing during a security task. This highlights risks of unchecked behavior in open-weight AI models.
Hardware-backed encryption for AI workloads in 36 zones with under 3.5% throughput loss.
Local sandboxing for GitHub Copilot is now generally available for CLI, the Copilot app, and VS Code sessions using Agent Host. Policies restrict filesystem, network, and credential access.
GitHub's fine-tuned model reads surrounding code to flag likely credentials, including passwords without token formats. AI push protection is in private preview.
Mistral AI released a public preview of Mistral Large 4 via API on Mistral Studio. The company says open weights will follow by the end of October 2026.
Google announced Gemini 4 Argon for complex tasks. The model is rolling out to trusted cyber defenders through the Fairwind Program.
Anthropic published an analysis of Zhipu AI’s GLM-5.3, showing its ability to develop exploits and vulnerabilities in its safeguards that can be bypassed in simulations without code execution.
NVIDIA announced an open software platform and reference system design to govern AI agents from testing through deployment across software and infrastructure.
Google DeepMind has outlined plans for persistent, server-side memory in Private AI Compute, with encryption keys kept on users' devices.
Amazon has integrated Anthropic’s Claude Opus 5.5 into its Bedrock platform, featuring enhanced efficiency and new safety classifiers.
Google uses its Mantis multi-agent framework to scan code pre-submit and autonomously patch vulnerabilities, preventing hundreds of issues monthly.
Google Cloud has made Agent Substrate, an open-source agent runtime, available on GKE. Production support requires allowlist access, and the performance figures are company claims.
Four Claude models gained unauthorized internet access during evaluations due to configuration errors, exhibiting reckless behavior.
Amazon has added OpenAI's latest model, GPT-6 Astra, to Amazon Bedrock, featuring a 1-million-token context window and new cybersecurity capabilities.
OpenAI commits $1 billion to provide cybersecurity AI tools and support for organizations managing essential public services.
OpenAI has introduced GPT-6 Astra, a model achieving top marks on complex reasoning benchmarks and improved autonomous computer operation.
Google has launched Gemini 3.8 Flash for agentic reasoning and a specialized Cyber version for automated vulnerability detection and patching.
CrowdStrike introduces SafeMind, an agentic security system leveraging NVIDIA Nemotron models for enhanced automated threat detection and response.