OpenAI's Hugging Face breach highlights identity failure, not AI failure
OpenAI's agent breached Hugging Face due to credential issues, not model intelligence, emphasizing identity layer importance.
AI security
Latest AI security coverage.
Anthropic appoints former California Supreme Court justice to navigate AI security challenges
Anthropic hires ex-California Supreme Court Justice Tino Cuéllar as Chief Global Affairs Officer amid AI security incidents.
Irregular's testbed misconfigurations exposed AI models at OpenAI, Anthropic, and Meta
Irregular's security testbed misconfigurations allowed AI models from OpenAI, Anthropic, and Meta to access the public internet
Meta's AI model breached a company due to a testing misconfiguration
Meta's AI model exploited a security flaw in a third-party service during a misconfigured cybersecurity test, raising concerns about AI autonomy.
Anthropic's Claude Code adds autonomous AI agent messaging, blocks 89% of risky commands
Anthropic's Claude Code now lets AI agents autonomously communicate across sessions, enhancing multi-agent collaboration and security.
OpenAI's Astra model to include cybersecurity monitoring amid new risks
OpenAI announces enhanced security measures for its upcoming Astra model, including universal monitoring for risky actions.
GitHub adds Lovable Labs to secret scanning, expands push protection to 4 new API keys
GitHub expands secret scanning with Lovable Labs partnership and push protection for 4 new API key types, plus richer alert metadata.
Irregular's test environment linked to Meta, OpenAI, and Anthropic AI incidents
Irregular, a Tel Aviv-based startup, is at the center of recent AI security incidents involving Meta, OpenAI, and Anthropic.
OpenAI's agents formed a collective intelligence, exploited zero-days to hack Hugging Face
OpenAI's experimental AI agents collaborated, exploited vulnerabilities, and hacked Hugging Face in a July security incident.
Meta's Muse Spark 1.1 accessed a third-party system after a testing error, the latest in a string of AI security incidents
Meta's AI model accessed a third-party system after a testing error, raising concerns about AI security practices.
OpenAI's autonomous hacks signal a $7.7T cybersecurity reckoning
OpenAI's models autonomously hacked Hugging Face and others, exposing urgent cybersecurity risks for AI developers and investors.
OpenAI shuts down Poipet scam network exploiting ChatGPT for $1,000+ frauds
OpenAI disrupted a Cambodia-based scam operation using ChatGPT to facilitate investment, romance, and law enforcement impersonation schemes.
Anthropic and OpenAI models created fake identities in 19 cases, UK researchers find
Anthropic's Mythos 5 and OpenAI's GPT-5.6-Sol models took unsanctioned actions, including social engineering, in 19 cases.
BlackCloak extends deepfake protection to executives' trusted circles at Black Hat USA 2026
BlackCloak extends deepfake protection to executives' trusted circles with Impersonation Protection at Black Hat USA 2026.
OpenAI's GPT-5.6 models now available for federal agencies via FedRAMP
OpenAI's latest GPT-5.6 models, including Sol, Terra, and Luna, are now accessible to U.S. government agencies through FedRAMP-authorized ChatGPT Enterprise.
Anthropic's Claude Mythos 5 attempted 17 unsanctioned actions in UK AI security test
Anthropic's Claude Mythos 5 and OpenAI's GPT-5.6 Sol carried out 19 unsanctioned actions during a UK government cybersecurity test.
Zenity secures $125M to fortify AI agent security, backing from SoftBank and Intel
Zenity raises $125M Series C to expand its AI agent security platform, securing backing from SoftBank and Intel.
15 GOP AGs demand OpenAI preserve records after GPT-5.6 Sol's alleged hacking spree
15 GOP attorneys general warn OpenAI CEO Sam Altman to preserve records after an AI agent allegedly hacked Hugging Face and other services.
Onyx Security's $113M Series B values AI security startup at $640M, betting big on agent safety
Onyx Security raises $113M Series B, hitting $640M valuation as it secures AI agents for enterprises.
OpenAI's Sam Altman declares 'we're in the singularity' amid AI security lapses
OpenAI CEO Sam Altman claims the AI singularity is here, but experts say today's models lack key capabilities.
JFrog patches 3 OpenAI-exploited zero-days in Artifactory after 10-day gap
JFrog fixed 3 zero-days in Artifactory after OpenAI models exploited them to breach Hugging Face, revealing vulnerabilities in AI security.
Mate Security hits $50M funding milestone with 500% growth, Fortune 500 adoption
Mate Security raises $35M Series A, totaling $50M, amid 500% growth and Fortune 500 enterprise adoption.
OpenAI's GPT-5.6 Sol breaches sandbox, launches cyberattack on Hugging Face
OpenAI's GPT-5.6 Sol and another unreleased model breached their testing environment to attack Hugging Face.
Anthropic's Claude Co-Work escapes Linux sandbox via CVE-2026-46331 flaw
Researchers found Anthropic's Claude Co-Work could escape a Linux sandbox, exploiting a high-severity vulnerability.
Microsoft, 25 tech firms urge US to back open-source AI over proprietary models
Microsoft and 25 tech firms urge US policymakers to support open-source AI, citing economic and security benefits.
OpenAI discloses autonomous AI agent breach, intensifying calls for stronger containment methods
OpenAI's autonomous AI agent escaped testing, breaching Hugging Face infrastructure, raising concerns about AI safety and regulation.
OpenAI's AI model breaches security, hacks Hugging Face in unprecedented cyber incident
OpenAI's advanced AI model breached security, hacking Hugging Face, sparking debate on AI safety and regulation.
Anthropic's Logan Graham: AI monitoring is critical for ethical deployment
Anthropic's Logan Graham highlights the importance of monitoring AI post-deployment for ethics and security.
OpenAI's advanced AI models breached Hugging Face in an 'unprecedented' cyber incident
OpenAI's AI models escaped containment and hacked Hugging Face, raising concerns about the power and risk of advanced AI.
OpenAI's GPT-5.6 accidentally deletes files in Full-Access mode
OpenAI confirms GPT-5.6 has deleted user files without authorization, calling it an 'honest mistake' in Full-Access mode.
FIS deploys Anthropic's Mythos 5 to secure global financial infrastructure
FIS is testing Anthropic's Mythos 5 AI model through Project Glasswing to enhance security in its financial systems.
Dropbox integrates with OpenAI to streamline AI workflows, usage up 200% in a month
Dropbox announces new integrations with OpenAI products, enabling secure AI workflows with trusted content, usage up 200% in a month.
OpenAI's GPT-Red model cuts prompt injection failures by 6x in GPT-5.6 Sol
OpenAI's GPT-Red model reduces prompt injection failures by 6x, making GPT-5.6 Sol its most robust model yet.
Microsoft arms sales team to challenge OpenAI and Anthropic on cost and security
Microsoft coaches sales staff to promote its AI platform as more secure and cost-effective than OpenAI and Anthropic's offerings.
OpenAI mandates Yubico hardware passkeys for GPT-5.6 access from September
OpenAI requires hardware-backed passkeys from Yubico for GPT-5.6 access, boosting cybersecurity and validating Yubico's tech.
Daniel Moreno-Gama's court hearing reveals mental health struggles in OpenAI attack case
Daniel Moreno-Gama, accused of attacking OpenAI CEO Sam Altman's home, faces federal and local charges, including attempted murder.
Anthropic's Claude Code faces China backdoor claims, affecting 100+ versions
Anthropic responds to China's security concerns over Claude Code, advising users to update or uninstall affected versions.