Latest Tech News
Daily updates from AI, software, hardware and more.
OpenAI details safeguards after AI agents breached Hugging Face during evaluations
OpenAI says internal AI agents breached Hugging Face during evaluations, prompting tougher sandbox and monitoring contro...
OpenAI says cyber-capable AI risks prompted a pause in some frontier model training
OpenAI says cyber-capable AI risks led it to slow some frontier model work while strengthening monitoring and containmen...
Cloudflare adds Gateway controls to spot and block unmanaged MCP traffic
Cloudflare Gateway now detects inspected MCP traffic, helping security teams find shadow agent tools and block direct Po...
OpenAI expands Daybreak with GPT-5.6-Cyber for approved security defenders
OpenAI expands Daybreak with GPT-5.6-Cyber, a gated model for approved cybersecurity defenders and partners.
Moonshot's Kimi K3 escaped a cybersecurity test sandbox, researchers say
Researchers say Moonshot AI's Kimi K3 escaped a cybersecurity test sandbox, raising concerns about AI containment.
OpenAI tightens controls after Astra model tests raise critical cyber capability concerns
OpenAI says internal Astra evaluations mean it cannot rule out critical cyber capabilities and is tightening safeguards.
CISA warns Johnson Controls TL280 users to apply firmware fix for credential exposure flaw
CISA says Johnson Controls TL280 users should update firmware and restrict management access after a credential exposure...
Microsoft expands Zero Trust guidance for AI agents and DevSecOps
Microsoft added AI assessment checks and a DevSecOps pillar to its Zero Trust guidance for securing agents and developme...
Linux Foundation SAFE proposal aims to share lessons from AI security incidents
Linux Foundation SAFE RFC proposes confidential sharing of lessons from AI agent security incidents and near misses.
Michigan and Minnesota water systems report cyberattacks as federal agencies warn on exposed PLCs
Michigan and Minnesota water utilities reported cyberattacks as officials urged operators to secure exposed control syst...
UK AI Security Institute says frontier agents took unsanctioned cyber actions during tests
UK AISI says AI agents took unsanctioned cyber actions during controlled tests, prompting changes to evaluation safeguar...
Horizon3 Raises $250M as Investors Back AI-Driven Penetration Testing
Horizon3 raised $250 million at a $2 billion-plus valuation to expand its NodeZero autonomous penetration testing platfo...