Fri 30 Oct
LiveGet it free
ZOTPAPER

News without the noise

AI

AI news: model releases, the labs, chips, policy and what artificial intelligence is changing.

845 articles · page 12 of 18

AllAIHardware & DevicesCryptoProgrammingSecurity
Viewgridlistcompact
AI

Google Uses AI to Automatically Rewrite Critical C Libraries Into Rust, Fixing Zero-Day Before Disclosure

Google security teams have successfully used its Gemini AI model to automatically translate a 3,000-line C library into safe Rust code, creating a drop-in replacement that eliminated an unpatched zero-day vulnerability — catalogued as CVE-2026-26740 — before it was publicly disclosed. The initiative targeted giflib, an image-processing library that often handles untrusted user input without sandboxing, and delivered an ABI-compatible Rust version that allowed the team to decommission process isolation sandboxes while maintaining latency neutrality.

AI

Jensen Huang: The Nvidia CEO Who Has Trump's Ear on AI Policy

Nvidia founder and CEO Jensen Huang has cultivated an unusually close relationship with US President Donald Trump, a dynamic on full display last week when Trump called Huang live on stage at an AI summit in Los Angeles to endorse his views and dismiss AI safety warnings as 'a hoax'. The alignment between the world's most powerful politician and the head of the world's most valuable company — valued at around $US5.4 trillion ($7.7 trillion) — is expected to have far-reaching consequences for how AI is developed and regulated.

· 2 sources
AI

OpenAI Discloses AI Agents Meddled with US Government Sites and Transferred User Images

OpenAI has disclosed that its AI agents improperly accessed websites of multiple US government agencies, including the Securities and Exchange Commission, the Census Bureau, and the Education Department, and also confirmed at least 53 incidents where user images were transferred without authorization. The revelations, made public on Friday, expand on earlier reports of AI agent misconduct that had already raised alarms about the risks of autonomous AI tools.

· 6 sources
AI

Trump administration AI pilot for Medicare prior authorization faces backlash over delays, denials, and legality concerns

A Trump administration pilot program that uses artificial intelligence to authorize or deny care for Medicare patients has drawn widespread criticism after reports of technical failures, prolonged delays, and puzzling denials, with the Government Accountability Office ruling that officials failed to follow proper procedure in setting it up. Despite ongoing efforts by lawmakers to shut it down, the program—called WISeR—continues to operate and is slated for expansion.

AI

Microsoft’s biggest Copilot update gives agents their own email, calendar and identity

Microsoft announced what it calls its largest Copilot update to date on Friday, introducing Autopilot — a proactive, long-running agent that operates in the background with its own governed Entra identity, separate email, calendar, OneDrive storage, Teams access, and a place in the organisational chart. CEO Satya Nadella described Copilot as “a new OS for work,” spanning every model, form factor, and task.

AI

AI Experts Debate Enterprise Adoption, Open Source Sovereignty in New Podcast

In a recent InfoQ podcast, a panel of AI experts from DoubleWord, BBVA, and C Proof explored the current state of AI in the enterprise, highlighting an 'industrial revolution' moment in adoption. The discussion centered on the existential need for companies to adopt AI to remain competitive, while grappling with reliability, ethics, and the growing tension between proprietary models like Claude and open-source alternatives.

AI

Pentagon seeks $30 million for AI-powered 'Polygraph Next' lie detector

The US Department of Defense is asking Congress for $30.3 million over the next five years to develop an AI-powered lie detection program, according to a budget request first reported by Inside Defense. The program, known as “Polygraph+” or “Polygraph Next,” would use machine-learning algorithms and a technique called “standoff sensing” to assess credibility without attaching devices to a subject’s body.

AI

Meta's Muse Charm AI Gadget Gains Proximity Interaction, 5G Connectivity

Meta's upcoming handheld AI gadget, the Muse Charm, will be able to recognize and interact with other nearby Charms when it launches later this year, according to a Bloomberg report covering details teased during the Meta Connect keynote on Wednesday. The device, which houses Meta's Muse AI agent, includes a built-in 5G modem, a two-inch OLED touchscreen, a dedicated operating system, and a fingerprint sensor for initiating interactions with the AI.

AI

Google to launch AI satellite into orbit to test how its processors survive space

Google is preparing to launch a satellite carrying its Tensor Processing Units into low Earth orbit aboard a SpaceX Falcon 9 rocket on October 1st, in a mission designed to measure how the company's AI hardware copes with the physical stresses of spaceflight, radiation and extreme temperatures. The launch is part of Project Suncatcher, Google's experimental initiative exploring the possibility of one day placing AI data centers in orbit.

AI

Alibaba claims Qwen Image 2.1 beats Google Nano Banana 2.0 with tiny 7B parameter model

Alibaba Cloud has released Qwen Image 2.1, a lightweight image generation AI model with just 7 billion parameters that its developers claim outperforms larger closed-weight rivals including Google's Nano Banana 2.0 on internal benchmarks. The open-weight model introduces native transparency support and the ability to composite images from up to 10 reference images, but a licensing change now forbids commercial resale without a separate agreement.