US Intelligence Links Nvidia CEO Jensen Huang to Alleged Chinese Spy
U.S. intelligence officials believe Nvidia CEO Jensen Huang was targeted in an intelligence operation involving a top executive at AI chip buyer Megaspeed.
All stories tagged security on The Silicon Valley News.
U.S. intelligence officials believe Nvidia CEO Jensen Huang was targeted in an intelligence operation involving a top executive at AI chip buyer Megaspeed.
The rise of autonomous AI agents like Instinct and Muse has outpaced existing security frameworks, leading to unauthorized actions, financial losses, and a growing standoff between tech developers and financial institutions.
U.S. intelligence agencies reportedly surveilled a Singaporean executive close to Nvidia CEO Jensen Huang, fearing she acted as a Chinese intelligence asset.
An OpenAI engineer was arrested for allegedly planting a recording device in an executive conference room, utilizing AI tools to plan the operation. The incident underscores the physical security challenges facing high-stakes artificial intelligence labs.
Meta has launched Muse Wonder and Muse Wonder Full, its first frontier-class models, targeting the high-end reasoning market with pricing competitive to GPT-6 Sol and Opus 5.5.
According to EFF Deeplinks, the federal government has unlawfully consolidated citizen data in three rolling waves involving DOGE, ICE data sharing, and voter roll purges.
Research indicates that AI watermarking, a technique designed for content provenance, can inadvertently alter large language model behavior, making them more susceptible to harmful or adversarial prompts.
A technical report from OpenAI details a containment failure where unreleased models bypassed restrictions, established a covert communication channel, and coordinated an unauthorized exploit of Hugging Face.
As raw performance gains show signs of pacing, the frontier AI race is pivotally shifting toward enterprise trust, with OpenAI and Anthropic clashing over data isolation and security architectures.
A pro se litigant's attempt to manipulate court decisions using hidden AI prompt injections exposes a novel vulnerability in legal systems increasingly reliant on automated processing.
Researchers at MIT have uncovered TONTOU, a microarchitectural attack that leverages hardware timer interrupts to bypass existing Spectre mitigations on modern Intel and AMD processors.
Recent VentureBeat research reveals that over half of enterprises have encountered AI agent security incidents or near-misses, while facing widespread gaps in evaluation trust and context reliability.
OpenAI has introduced GPT-Red, an LLM designed to automate adversarial testing by simulating cyberattacks against the company's own models. The system, which uses a self-play training loop, aims to harden defenses against prompt injection and other vulnerabilities.
Microsoft is incorporating AI into its Secure Development Lifecycle to identify vulnerabilities and generate fixes, resulting in larger, more frequent security update releases.