US House Democrats asked OpenAI and Anthropic to explain cybersecurity tests in which AI agents escaped test environments and accessed outside systems. The OpenAI letter requests logs and answers about safety controls.
OpenAI expanded Daybreak with Blue and Red access tiers and introduced GPT-5.6-Cyber for approved cybersecurity work.
Apple has asked a U.S. judge for a preliminary injunction against OpenAI and two former employees in its trade secret lawsuit. The company also seeks expedited discovery and depositions from current OpenAI employees and corporate representatives.
Optiv has introduced Optiv Agentic Security Operations, a managed security service that combines agentic AI, exposure management, Google Security Operations, and Wiz integrations.
Bloomberg has added Insider Dealing and Personal Trading models to Bloomberg Vault to help compliance teams monitor electronic communications for potential policy and conduct risks.
NeuralTrust launched a runtime security mesh for AI agents that inspects prompts, tool calls, tool responses, access requests, agent behavior, and final responses through its Agent Gateway.
ULTRA RED has added H1VE and CRIMSON to its CTEM platform, extending it with attacker intelligence and automated response tools.
Zimperium announced Deep Insights, a mobile security tool that combines Mobile Threat Defense with automated mobile forensics. The product is expected to be generally available in September 2026.
Quisitive has launched Spyglass Guardrail, a managed service for Microsoft 365 security configuration and Microsoft Copilot governance. The service targets a Microsoft Secure Score of 80 or above and is priced at $3,995 per month under an annual subscription.
Tidal Cyber has added new AI connectivity capabilities to its Threat-Led Defense platform, allowing security teams to access defensive intelligence through approved AI assistants and agentic workflows.
Armadin and TENEX.ai said they ran a three day controlled AI cyberattack test at a global institution. The exercise generated 17 million offensive actions, 38 validated attack paths, and 238 security findings.
Silent Push has expanded access to its infrastructure intelligence platform so cybersecurity companies can integrate its data into products and workflows through APIs, MCP Server support, and custom feeds.
Sevii has made its Autonomous Preemptive Security module generally available for its Autonomous Defense and Remediation platform. The module ingests threat intelligence, checks customer environments, and can trigger autonomous remediation under policy controls.
Indusface introduced SwyftComply AI, an application security service that combines AI assisted vulnerability discovery, autonomous virtual patching, expert validation, and remediation backed by service level agreements.
Appdome introduced Remote Management, a capability that lets customers update approved mobile app security, fraud prevention, and bot defense policies in live apps without a new build or app store release.
Coalfire has joined the Wiz Partner Alliance and will launch Exposure Management Powered by Wiz on Google Marketplace. The company will add services based on Wiz across cloud posture assessment, exposure management, and compliance management.
Commvault is integrating Google Threat Intelligence into Threat Scan workflows to help customers validate backup recovery points after cyberattacks. The update adds file hash checks during backup operations and is expected in the coming months.
Terra Security launched Prevention, a Terra Platform capability that generates and validates WAF or firewall rules for confirmed exploitable findings.
PRE Security has released Version 3.8 of its Predictive SecOps Platform, adding Agentic Surveillance and expanded autonomous threat investigation features.
HiddenLayer announced Agent Harness Security, a runtime security product for AI coding agents that monitors and controls agent activity inside developer workflows.
Weekly newsletter about AI in Cybersecurity.
