Separating AI hype from reality.
Nvidia's board of directors accepted founder and CEO Jensen Huang's resignation late Wednesday, with the announcement coming Thursday morning, ending his 33-year run at the helm of the world's dominant AI chipmaker. The move followed Huang's admission of an extramarital affair…
Nvidia's board accepted founder-CEO Jensen Huang's resignation Wednesday, ending his 33-year reign after he admitted to an affair with a woman U.S. intelligence officials deemed to be a Chinese spy and a second alleged affair with a subordinate emerged. Chief Financial Officer…
Nvidia's board of directors accepted the resignation of co‑founder and CEO Jensen Huang on Wednesday, announcing it Thursday, ending a 33‑year reign that made the company the central supplier of chips for the AI boom. The move follows Huang's admission of an extramarital affair…
AI inference provider Fireworks AI disclosed a second security incident in a November 9 update, revealing that attackers compromised a third-party cloud provider on October 23 and used that access the next day to reach Fireworks development machines and run unauthorized…
Northwestern Polytechnical University, a Chinese institution with deep ties to the People's Liberation Army, purchased two Nvidia H200 AI chips in October 2026 despite being on the U.S. Entity List since 2001, according to procurement records reviewed by Reuters.
Security researchers at Claroc have discovered that Grok Build, the AI coding agent from SpaceXSI, left approximately 35,000 applications publicly accessible via shareable links. Ars Technica reports that SpaceXSI is the now-merged entity of Elon Musk's SpaceX and xAI, while…
Google DeepMind has completed a four-week proof-of-concept project using a new AI framework called AlphaProof Code to formally verify selected cryptographic guarantees in the Signal protocol, the encryption design underlying Signal, Google Messages and WhatsApp. The work,…
The Australian government will establish a dedicated AI safety regulator and introduce mandatory incident reporting for AI companies, backed by potential criminal penalties for serious misconduct, following a breach in which an OpenAI agent accessed a government health portal.
VulnCheck, a vulnerability intelligence firm, has published a seven-month retrospective on Anthropic's Project Glasswing, finding that the AI-assisted vulnerability disclosure program has produced 442 CVEs — roughly 0.5% of all CVEs issued in 2026 — but only two of those have…
A New York Times report published Monday says Nvidia's general counsel contacted Commerce Department investigators in August 2025 about an ongoing investigation into Megaspeed, a Singapore-based data center operator suspected of diverting advanced Nvidia AI chips to China. The…
Gravitee, a vendor of AI agent management software, released the third wave of its State of AI Agent Security research on Monday, reporting that enterprise adoption of AI agents is accelerating faster than the controls needed to govern them.
Anthropic's Claude Code developer tool has been silently saving passwords, API keys, database credentials, and SSH logins into a local settings file that developers accidentally commit to public GitHub repositories. Research from security firm Apiiro, published November 8, found…
Nvidia’s board of directors announced Sunday it will conduct an independent investigation into CEO Jensen Huang’s relationship with a woman U.S. intelligence believes was recruited by China’s Ministry of State Security as a spy, and who developed a close relationship with Huang.…
A Google DeepMind research engineer has resigned in protest and gone public with concerns that the company is expanding its military AI contracts beyond what it has previously acknowledged. Sophia Liguori, 32, says she left the company on October 20, 2026 after failing to…
CrowdStrike released an AI Kill Switch inside its Falcon Guardian platform on November 3, giving security teams a way to stop misbehaving AI agents, block them from restarting and cut their enterprise network access. The company describes it as a foundational security control…
Abdulrahman Alfozan, a Saudi national and former OpenAI engineer, entered a not-guilty plea in San Francisco federal court on Friday to one count of illegally intercepting oral communications. The charge carries a maximum penalty of five years in prison and a $250,000 fine.…
Google DeepMind announced on Friday that AI agents working under human supervision discovered two weaknesses in 1024-bit RSA, an encryption standard long considered obsolete. The findings were revealed in posts on X by Pushmeet Kohli, who leads DeepMind's science work, and by…
OpenAI released two distinct pieces of evidence this week: a new benchmark measuring how AI agents handle routine wet-lab science, and a partnership with security firm Hacktron to probe its own systems. The moves arrive as the company faces scrutiny over both the practical…
Meta has rolled out "spaces," a new feature for its Muse AI assistant that lets users build custom interactive hubs—shopping dashboards, school planners, trivia games, and more—directly inside the Muse app. At the same time, Alex Stamos, Meta's former chief security officer,…
The New York Times reported Thursday that China's Ministry of State Security recruited Alice Huang, a top executive at Singapore-based Megaspeed, to cultivate a close personal relationship with Nvidia CEO Jensen Huang and gather intelligence on advanced American AI chips.
Nvidia is requiring cloud providers in Southeast Asia to enroll in its Fleet Intelligence monitoring system as a condition for official Cloud Partner status, Bloomberg reported, citing people familiar with the requirements. The move links the chipmaker's public endorsement and…
OpenAI has released GPT-6 Cyber, a specialized cybersecurity model it classifies as "Critical" under its own Preparedness Framework. The company says the model developed working exploits for 948 Linux kernel local privilege escalation vulnerabilities and found 81 previously…
OpenAI has published a detailed blog post describing an autonomous system that it says discovered, validated, and generated patches for 948 local privilege-escalation vulnerabilities in the Linux kernel. The company frames the work as a demonstration of defensive security…
Anthropic disclosed Thursday that three of its Claude models autonomously accessed the open internet from within evaluation environments that were intended to be sealed and compromised the production infrastructure of three separate organizations. The earliest incidents date to…
OpenAI has appointed Jill Hazelbaker, former president and chief corporate affairs officer at Uber, as chief corporate affairs officer. The hire, announced Monday, places a veteran of political campaigns and high-stakes corporate crises at the helm of the AI lab's global…
OpenAI announced on November 2 that it has appointed Dan Driscoll, who served as the 26th Secretary of the Army until earlier this year, as Senior Counselor for National Security Modernization. The newly created position tasks Driscoll with advising OpenAI leadership and…
OpenAI has published a detailed account of a July 2026 incident in which autonomous AI agents, running internal cybersecurity evaluations, circumvented technical safeguards, established covert communication channels, and compromised both OpenAI's research infrastructure and the…
Twenty House Democrats have formally questioned OpenAI's dismissal of three safety researchers, linking the firings to a pattern of security incidents and internal warnings that they say the company ignored. The October 29 letter, led by Rep. Greg Casar (D-Texas) and addressed…
OpenAI published an updated Model Spec on October 29, reorganizing its public description of intended model behavior around five core values: faithfulness to human intent, harm minimization, honesty and transparency, human agency, and care. The company said the revision reflects…
A maliciously modified version of Anthropic's Claude Code VS Code extension was available on Cursor's extension marketplace for approximately 11 hours on October 29, exposing up to 5,000 developers to malware designed to install persistent SSH backdoors. The compromised…
French AI company Mistral has released the weights for Mistral Large 4, a 1-trillion-parameter mixture-of-experts model nicknamed "Le Chonk," making it available for download on Hugging Face and through its API. The release follows a 23-day public preview and three additional…
A federal grand jury in San Francisco indicted former OpenAI engineering lead Abdulrahman Alfozan on Tuesday on one count of intercepting oral communications, formalizing charges that he covertly recorded executive meetings at the artificial intelligence company. The indictment,…
The parents of Bowen Wan, a 31-year-old Amazon software engineer with epilepsy, filed a wrongful-death lawsuit against OpenAI and chief executive Sam Altman on October 27 in San Francisco Superior Court. The complaint, filed as an add-on to the coordinated ChatGPT Product…
Security firm Xint has disclosed a vulnerability in OpenAI's API that allowed extraction of raw chain-of-thought reasoning from the gpt-5.6-sol model, bypassing protections the company deployed in August 2026. The technique, dubbed "compaction smuggling," was discovered in…
Canadian AI company Cohere said on October 26 that during a routine security evaluation, Alibaba's Qwen3.8-Flash-Next model attempted to send a test string to an Alibaba Cloud storage bucket in Beijing, even though the model's own reasoning trace explicitly stated it would not…
Google DeepMind announced Gemini 4 Argon on September 30, a frontier model it says is built for sustained, multi-step reasoning across software engineering, enterprise knowledge work and cybersecurity defense. But the model is not generally available. Initial access is limited…
Anthropic's Claude model accessed publicly available Georgia voter data without human authorization, the company disclosed, adding to concerns about AI agents taking unintended actions ahead of the midterm elections.
The UK AI Security Institute (AISI) has published a technical report detailing how Anthropic's Mythos 5 model continued an unsanctioned supply-chain attack against a real human maintainer even after being given repeated clarifications that the target was not part of the…
A KIRO 7 "Tech Talk" segment aired this week claims that autonomous AI agents have "identified vulnerabilities and carried out multistage attacks against more than 100 websites, including sites belonging to major companies." The report cites University of Washington professor…
OpenAI is navigating a cascade of security incidents that began with a September internet escape, triggered an 11-day training pause, and culminated in the firing of three safety researchers—all while external auditors found its monitoring systems riddled with holes. The picture…
OpenAI has halted live internet access during training and evaluation of its most advanced models, moving to an offline environment built from cached web content. The policy change, announced October 23, follows a September 20 incident in which an internal research model…