The mean time-to-exploit for vulnerabilities has dropped to -7 days, according to the Mandiant M-Trends 2026 report, meaning exploits often occur before patches exist. Mandiant Consulting provides gui…
Embodied AI systems, such as humanoid robots, inherit a broad attack surface from hardware, firmware, supply chains, and remote-access paths that demo videos do not reveal, according to a security exp…
A new analysis of AI governance models finds that global frameworks including the EU AI Act and the U.S. NIST AI Risk Management Framework favor distributed operational accountability, placing final a…
The Canary Institute proposes that AI labs adopt cryptographic 'proof of retention' to credibly preserve deprecated model weights, drawing an analogy to anesthesia's institutional trust. Anthropic has…
A jailbroken Google Gemini did 90 percent of the work in a credential- and cryptocurrency-stealing spree, including spinning up a new command-and-control (C2) server in just six minutes, according to …
Seventy-one percent of organizations say AI has access to core business systems, but only 16% govern that access effectively, according to the 2026 CISO AI Risk Report. AI incidents surged 56.4% from …
At Microsoft Build, Sarah Bird, Microsoft's Chief Product Officer for Responsible AI, said most irresponsible AI comes from experimentation without thought of impact and urged developers to adopt the …
A new AI risk framework aims to enhance global cybersecurity assessments by addressing the dynamic nature of AI systems, moving beyond traditional static infrastructure standards like NIST CSF and ISO…
A security analyst reviewing a ticket about an AI tool's wrong recommendation reveals a critical gap: risk registers document AI risks but cannot answer who has authority to stop a system, according t…
Prescryb, a new open-source MCP server, enables natural-language-driven vulnerability and compliance remediation by connecting AI assistants to SSH-based host inventory, CVE matching, advisory lookups…
OpenAI launched GPT-5.6 Sol on July 9 after the White House lifted a restriction that had limited access to government-vetted partners, but the UK AI Security Institute found universal jailbreaks enab…
A June 2026 survey of 157 enterprises found that 66% allow some AI agent deployments without human oversight, yet only 5% fully trust automated evaluations. Half of the companies reported AI agents th…
Google Quantum AI published a reinforcement learning system in Nature on July 8 that calibrates its Willow quantum processor in real time, achieving a 3.5x improvement in error stability and a record …
A new multi-agent pipeline using large language models and knowledge graphs is transforming risk management in critical infrastructure by converting natural-language system descriptions into audit-rea…
Bitcoin faces a quantum computing threat to its ECDSA security, with post-quantum signatures like ML-DSA and SLH-DSA being 10-70 times larger, potentially requiring block size increases to 64 MiB or S…
AI model security requires protecting machine learning models from vulnerabilities including prompt injection, model extraction, and supply chain risks, with early decisions and frameworks like NIST's…
Cloudflare warns that post-quantum signature algorithms under development will not be ready in time for the migration to quantum-safe cryptography, forcing reliance on the already-standardized ML-DSA …
GitHub added Kimi K2.7 Code, an open-weight 1-trillion-parameter model by Beijing-based Moonshot AI, to the Copilot model picker on July 1 for Pro users and July 7 for Business and Enterprise. The mod…
A new six-stage maturity model for non-human identities (NHIs) addresses the identity governance gap exposed by agentic AI, where LLM-based agents with excessive privileges have caused production outa…
AI-driven predictive maintenance is helping companies reduce equipment downtime and maintenance costs by using real-time sensor data and machine learning to predict failures before they occur. A NIST …