NVIDIA launches SkillSpector, a free open-source AI security scanner that detects vulnerabilities and malicious patterns in AI agent skills before installation, targeting a growing threat landscape where over 26% of skills contain security risks.
A shocking new survey finds two-thirds of office workers are secretly using unauthorized AI tools at work, with 88% feeding sensitive company data — including customer records and financial documents — into public chatbots, creating what experts warn is a massive and growing enterprise security liability.
Anthropic reverses a secret policy that would have covertly degraded Claude Fable 5's performance for researchers building competing AI models, now committing to transparent safeguards after fierce backlash warning the original approach threatened open-source AI development.
Sentry's new MCP server integrates directly into Cursor IDE and other AI-powered editors, enabling developers to query production bugs, run AI-driven root cause analysis via Seer, and apply fixes instantly without ever leaving their coding environment.
A viral GitHub repository called CL4R1T4S is exposing hidden system prompts from ChatGPT, Claude, Gemini, Grok, and other major AI platforms, amassing over 27,800 stars as it publicly reveals the secret instructions quietly shaping how these AI systems think and respond.
A German court holds Google liable for defamatory AI-generated search summaries after the tool falsely linked two publishers to scams, marking a landmark first in AI legal accountability and potentially opening the floodgates for global lawsuits against AI firms.
Anthropic unveils two major AI policy frameworks backed by $350 million in commitments, as CEO Dario Amodei urges urgent regulation ahead of a near-trillion dollar IPO, raising questions about whether the move is genuine risk mitigation or strategic PR.
Bugbot just received a massive upgrade making it over 3x faster, 22% cheaper, and 10% more effective at catching bugs, with 90% of runs now completing in under three minutes thanks to harness upgrades and the new Composer 2.5 model integration.
Microsoft blocks internal employee access to Anthropic's Claude Fable 5 after the new AI model's 30-day data retention policy — and up to two years for flagged content — raises legal concerns over confidential data, while external GitHub Copilot and Foundry customers retain access as Microsoft's legal team evaluates the …
Databricks launches OpenSharing under the Linux Foundation as a vendor-neutral open standard for secure data and AI collaboration, backed by major players including OpenAI, SAP, and Stripe, expanding cross-platform support for AI models, agent skills, and unstructured data sharing across organizations.