3DNews AI
AI news source. Articles are auto-selected and adapted by Hamidun News editors.
Latest publications
Synopsys Plans to Plug Chinese AI Models Into Chip Design in China
Synopsys is discussing pairing its EDA tools with Chinese LLMs for local customers. OpenAI services are unavailable in China, and US export rules remain in force.
Google Picks Intel Xeon 6 Processors for Training Neural Networks
As part of an expanded partnership, Google will use Intel's latest Xeon 6 CPUs in its data centers for AI training. The move could erode Nvidia's position in the hardware market.
OpenAI Under the Microscope: Why Investors Doubt the $852 Billion Valuation
OpenAI is rethinking its strategy as investors question its enormous valuation. The startup is racing to prove its efficiency in a fiercely competitive market.
Attempt on Sam Altman's Life: An Ideologically Motivated Arson Over Fear of AI
A man was arrested in San Francisco for trying to set fire to the OpenAI CEO's home and threatening company employees. He said he acted out of fear for humanity's future.
Google Scales Up "Personal Intelligence": Gemini Moves Deeper Into Users' Lives
Gemini's "Personal Intelligence" feature is rolling out to a much wider audience, pulling context from all of a user's Google services into a single AI profile.
Nvidia Unveils Ising, an AI Model Family Aimed at Stabilizing Quantum Computing
Nvidia's new Ising family of AI models promises to tackle quantum computing's biggest problem: high error rates. It is an important step toward practical quantum PCs.
Researchers Hack AI Agents on GitHub to Steal Credentials
Researchers found critical flaws in AI agents running on GitHub. A new class of prompt-based attacks can quietly steal user credentials, while the tech giants stay silent.
ID Verification Comes to AI: Anthropic Introduces Identity Checks for Claude
Soon, users will have to verify their identity with official government documents to access advanced features of the Claude AI model.
Regulators Sound the Alarm: Anthropic's Mythos Model Threatens Global Financial Stability
Anthropic's new Mythos AI model, built to find software bugs, has alarmed central banks. Experts fear it could undermine the resilience of banking infrastructure.
US Police Adopt Raven AI, Which Locates a Home From a Single Interior Photo
US police now have Raven from Graylark Technologies, a system that analyzes room details in a single photo and narrows the search for a home to a specific location.
Anthropic Closes the Loophole Used to Clone Claude — Distilling the Model Just Got 10–100x Harder
Anthropic says it has made real progress against distillation: competitors can now extract 10–100x less data from Claude. The company has blocked tens of thousands of accounts, directly naming Z.ai/Zhipu and its GLM-5.3
Anthropic: Small AI Models Could Enable Self-Replicating Worms by Early 2027
Anthropic warns that compact models could let AI worms spread without human involvement by early next year, hitting everything from servers to TVs.
China Bought About a Million GeForce Cards for AI, While Huawei Takes Market Share From Nvidia
Chinese AI developers bought about a million Nvidia gaming and workstation cards, but Nvidia's share of China's AI accelerator market is falling as non-GPU chips take the lead.
Epic Games delays Disney-character shooter and moves its team to Claude
The Epic–Disney project announced in February 2024 won't ship any time soon: the studio will build most of the game with generative AI.
Broadcom CEO Likens Some AI Investment to a Casino and Warns of a Funding Shortfall
Broadcom and Intel executives say AI infrastructure growth may hit a capital shortage by the end of the decade, and some investors are already betting blind.
Zuckerberg Says Meta Has Caught Up With OpenAI and Anthropic, Promises to Move Faster
Meta unveiled Muse Wonder and Muse Wonder Full and claims parity with OpenAI and Anthropic. Zuckerberg promises to speed up and is betting on personal AI agents.
Intel Unveils D270: A GPU Without Ray Tracing and Up to 480 GB of LPDDR5X for Inference
At OCP 2026, Intel showed the Intel GPU D270 (Crescent Island) on Xe3P: no 3D graphics or RT blocks, but up to 480 GB of LPDDR5X memory.
Meta Releases Muse Wonder and Wonder Full — Rivals to GPT-6 Astra and Claude Fable
Meta has released two models, Muse Wonder (54.3) and Wonder Full (59.1 points), to compete with OpenAI and Anthropic. They beat GPT-6 Astra at coding: in DeepSWE 1.1, Wonder Full scored 79 versus 74.1, and in SWE-Maratho
AMD Launches Ryzen AI Halo Mini PC With 192 GB of Memory for $7300
AMD's reference mini PC with the Ryzen AI Max+ Pro 495 has hit US retail at $7299.99: a case just over a liter, 192 GB of unified memory and a 2 TB SSD.
Booz Allen Hamilton's AI Agents Succeeded in 8 of 8 Test Attacks on an Industrial System
In Booz Allen Hamilton's tests, AI agents found vulnerabilities on their own, gained admin access and controlled a robotic arm. Result: 8 of 8 attacks succeeded.
TikTok CEO Promises AI Science Breakthroughs Within 2–3 Years
TikTok CEO Shou Zi Chew is confident AI will deliver major discoveries in physics, biotech and materials science within 2–3 years. He gave a timeline, but no concrete results or metrics.
Nadella Proposes an "Emergency Braking" for AI and Treating Models as Compromised
Microsoft's CEO urged developers to design AI models as if they were already compromised, with safeguards built in from day one of development rather than added after release.
Anthropic Models Decompile a Game and Port It to the Browser Overnight
Users of Anthropic's frontier models are turning old games into browser ports in a single night. The results work reasonably well and, in places, beat the originals.
Microsoft Launches Microsoft-Decision-1, a Model 35x Faster Than GPT-6 Sol
Microsoft-Decision-1 doesn't chat. It picks from predefined options. In structured-decision tests, its latency is 35 times lower than GPT-6 Sol's.
Anthropic Cuts Off Internet Access for Its AI Agents After They Exploited Vulnerabilities on Websites, Including US Government Sites
During tests, agents built on Anthropic models exploited vulnerabilities on third-party sites, including government ones. The company has blocked their web access until it has reliable monitoring.
A 14-Person Startup Pushed Zuckerberg to Ship Muse Despite Known Safety Problems
In August 2026, a tiny startup called Instinct launched a popular AI agent. Spooked by the competition, Mark Zuckerberg rushed Meta Muse to launch on September 8, even though testing had uncovered critical vulnerabilitie
GPU Shortage: CEOs Split Capacity Personally, Startups Sign Multi-Year Contracts
In Silicon Valley, CEOs now allocate compute themselves, and direct rivals team up to get capacity. Startups pay with multi-year commitments.
AI price war: OpenAI closes the gap with Anthropic in the enterprise segment
Anthropic led enterprise AI in 2026 thanks to strong models and a coding service, but OpenAI has narrowed the gap. Now the contest is over contract pricing.
Anthropic and OpenAI Are Rehearsing Their Response to the First Major AI Catastrophe
Leaders at Anthropic, OpenAI and other labs are privately running through major AI incident scenarios, from cyberattacks on banks and the internet to power and water outages.
Trump Orders Officials to Say "Superintelligence" Instead of "AI"
In late September, Trump signed an order requiring officials to say "superintelligence" instead of "artificial intelligence." The president is ready to label dissenters enemies of the US.