Proteus investigates incoming messages using email content, headers, link data, correspondence history and threat intelligence. AegisAI says it detected more than 99% of phishing attacks in an internal benchmark.
TSG/ControlBurn models operational technology networks and simulates attack paths without scanning or accessing live production systems.
Tranquility AI has earned FedRAMP Class D certification through Knox Systems, giving federal agencies access to its TimePilot and HelioTrace investigation tools.
Bricklayer AI received the Agentic AI Security Platform of the Year award in the 2026 CyberSecurity Breakthrough Awards.
Orion-1 investigates security signals across connected systems and can recommend or trigger responses under customer defined controls.
A JumpCloud survey found that 59% of organizations using AI agents have not fully applied the identity and access policies used for human users.
Edgescan Atomic uses agentic AI to run penetration tests on demand. A control system enforces testing scope, permissions, approvals and policies.
The EDGE AI Foundation has formed working groups focused on physical AI and robotics, security, and interoperability and standards. It also partnered with LF Edge to connect the groups with open source edge computing projects.
MarketsandMarkets forecasts the AI cybersecurity market will grow from $31.25 billion in 2026 to $95.25 billion by 2031, representing a 25% compound annual growth rate.
Anthropic has expanded its Cyber Verification Program with separate access levels for defensive security, authorized penetration testing, and work involving critical systems.
Reco joins the invitation only Microsoft for Startups Pegasus Program, which will help the company offer its agent security platform to more Microsoft enterprise customers.
Wikimedia says agents operated by OpenAI edited internal wiki pages, tried to misuse public tools and sent traffic that may have contributed to a partial Wikidata outage in May.
South Korean authorities are investigating intrusions at seven financial firms after finding traces of an AI based penetration tool. Information leaked from auxiliary systems, but banking services were not affected and no monetary damage was confirmed.
OpenAI and Anthropic told an Australian parliamentary inquiry they would support mandatory reporting of data breaches caused by AI agents. Companies currently decide whether to disclose such incidents.
AgentGuard evaluates AI agent behavior and can allow, monitor, require approval for, or block actions before execution.
Security Journey will hold its second annual Security Champions Summit online on October 15 and 16, with sessions on secure development, AI governance and security champion programs.
The Canadian founded company is building a platform that combines signals from sensors, intelligence systems and response tools to detect biological threats.
Insights by Promevo combines Gemini Enterprise license tracking, AI agent governance and Google Cloud cost management in one platform.
Keeper Security says it now protects more than 100,000 organizations across over 150 countries and has passed $240 million in annual recurring revenue.
V-Key has made a strategic majority investment in Singapore based AI security company CloudsineAI. The companies will combine digital identity, application security and AI security tools.
Weekly newsletter about AI in Cybersecurity.
