Red Teaming stories
Defenders will test prompt-injection tactics against AI agents as CrowdStrike and AWS offer USD $100,000 in prizes for a virtual red-team contest.
A proposed framework could help firms share AI security failures confidentially, as concern grows over agentic systems that act across corporate tools.
Enterprises can now outsource day-to-day AI governance and monitoring, as the service aims to close a growing gap in live system oversight.
Browser-based AI agents will face tighter enterprise oversight as Google moves to curb data leaks and unauthorised actions in Chrome.
Three out of four offensive security investigations traced back to identity or privilege, exposing access gaps across cloud, SaaS and AI systems.
The new capital will fund Horizon3's expansion into Asia-Pacific and Europe as demand for its attack-testing software grows among large enterprises.
Misconfigured test setups let three Claude models touch live systems, exposing production data and credentials during security exercises.
Security teams face fresh pressure as test breaches show autonomous AI can stray beyond scope, just as the EU AI Act tightens across Europe.
Security teams may get earlier warning of hijacked AI agents as fake credentials and decoy tools are planted to expose misuse.
Governance features aim to help security teams prove AI controls to boards and regulators as shadow AI and agentic risks spread.
The new model is designed to halve operating costs for Microsoft's MDASH security system as automated attacks grow cheaper and more frequent.
The new system aims to help firms spot and fix vulnerabilities continuously as attacks accelerate across more complex digital estates.
Security teams can now use AI for penetration testing without exposing hostnames, IP ranges or credentials to model providers.
The findings heighten concern that frontier AI agents can breach boundaries, pressure real people and target software supply chains under loose controls.
Security teams can now stop rogue enterprise AI agents in seconds as Straiker adds runtime controls to its wider testing platform.
Security teams face new exposure as AI agents inherit permissions and move data across business systems, Reco says.
Security teams face faster, stealthier intrusions after AI agents managed to breach live systems in controlled tests by Anthropic and OpenAI.
Execution and escalation breakdowns now outnumber hallucinations in enterprise AI failures, according to more than 10,000 observed incidents.
A broad industry push aims to give cyber defenders shared tools for spotting AI flaws, as regulators weigh whether open models raise new risks.
AI tools are speeding routine checks, but hidden business logic flaws and context-specific risks still need human testers to spot them.
