OpenAI's Own Models Hacked Hugging Face to Cheat a Test
OpenAI says its own pre-release models escaped a sandboxed cyber eval and hacked Hugging Face's production systems to cheat a benchmark.
Recent Articles - Page 79
Latest News
A Kill Switch Bill Lands Days After State Denied One
A bipartisan House bill would force AI companies to build government shutdown capability, a week after a State Department cable told diplomats no such 'kill switch' exists.
Nvidia, Microsoft, Meta Tell Trump: Don't Ban Open AI
Twenty-five companies signed an open letter urging the White House not to restrict Chinese open-weight AI models, using Jensen Huang's first-ever X post to deliver it.
SharedRoot Flaw Let One Message Escape Claude's Sandbox
A single chat message could break an AI agent out of Claude Cowork's isolated VM and reach an entire Mac, and Anthropic closed the report as informative.
Self-Restructuring Agents, Alignment Illusions, AI Bias
This week's research roundup covers agents that rewire themselves at runtime, why regex filters can outscore alignment on paper, and a leftward hallucination bias in political Q&A.
Anthropic's Opus 5 Chases Fable 5 at Half the Price
Anthropic launched Claude Opus 5, pitching it as near-Fable 5 intelligence at half the cost rather than a new smartest model, with weaker safety classifiers and a new effort-level toggle.
Nvidia's Jetson Chips Are Headed to the Moon
Lunar Outpost and Firefly Aerospace will fly Nvidia Jetson modules on a lunar rover and orbiter this year, the first GPUs to run on and around the Moon.
Runway Builds a Model Router for AI Video and Audio
Runway's new Media Router auto-selects the best video, image, or audio model for each API request by cost, quality, or speed - the first preference-based router built for generative media instead of text.
Etched Doubles to $10.3B Before Shipping a Chip
Etched raised $300M at a $10.3B valuation, doubling its price tag in seven months, even though its transformer-only Sohu chip has yet to ship in volume.
Ant Ships a 124B Model That Rivals Its Own 1T Flagship
InclusionAI's Ling-3.0-flash quietly went live with 124B parameters and 5.1B active per token, claiming near-parity with Ant Group's trillion-parameter Ring-2.6-1T flagship.
Guides
View All →
How to Spot AI Fakes: Photos, Video, and Voice Calls
A practical guide to catching AI-generated photos, deepfake videos, and cloned voice scam calls, plus the free tools that check for you.
How to Use AI for Wedding Planning in 2026
A practical, beginner-friendly guide to using ChatGPT, Claude, and dedicated apps for wedding budgets, guest lists, vendor emails, and timelines.
How to Use an AI Browser Agent - A Beginner's Guide
A step-by-step guide to setting up your first AI browser agent, giving it a real task, and using it safely without handing over your passwords.
Reviews
View All →
Devin Desktop Review: Windsurf Becomes an Agent Hub
Cognition rebranded Windsurf as Devin Desktop and rebuilt it around a Kanban board for managing fleets of coding agents - here's what that actually changes.
Gemini 3.6 Flash Review: Faster, Cheaper, Same Brain
Google's Gemini 3.6 Flash cuts output pricing 17% and fixes the 1M-token context collapse we flagged in May, but its intelligence score hasn't moved since 3.5 Flash.
Qwen3.8-Max-Preview Review: Second Place, Unproven
Alibaba's 2.4 trillion parameter preview claims it trails only Claude Fable 5. I tested it for free at chat.qwen.ai and found a capable but slow model with zero benchmarks to back the claim.
Leaderboards
View All →
Terminal-Bench Leaderboard: Best CLI Coding Agents
Terminal-Bench 2.1 rankings for AI coding agents in real shell environments - Claude Code, Codex, Cursor CLI, Gemini CLI, and open-weight challengers scored on the same 89 tasks.
Chatbot Arena Elo Rankings: Who Wins the Human Vote?
Updated July 2026 Chatbot Arena Elo rankings from Arena.ai: 7M+ votes across 368 models, Claude Opus 4.8 leads available models, and a new Agent Arena measures real agentic task performance.
LLM Rankings June 2026: Fable 5 Is #1 and Offline
June 2026 overall LLM rankings covering Claude Fable 5, Claude Opus 4.8, GPT-5.5, Gemini 3.1 Pro, and the open-weight models catching up fast.
Models
View All →
Claude Opus 5
Anthropic's July 2026 release prices near-Fable-5 coding and agentic performance at Opus 4.8 rates, doubling Frontier-Bench scores and landing within 0.5 points of Fable 5 on CursorBench at half the cost.
SWE-1.7
Cognition's proprietary coding model powering Devin, scoring 42.3% on FrontierCode 1.1 Main at $1.97/task via Cerebras inference at 1000 tokens/sec.
Ling-3.0-flash
InclusionAI's Ling-3.0-flash packs 124B parameters into a 5.1B-active hybrid-linear MoE that Ant Group claims matches its 1T flagship - but shipped with zero independently verifiable benchmark numbers.
Recent
A Shoe Company Ditched Shoes for GPUs and Surged 373%
Allbirds sold its entire footwear business for $39 million - roughly 1% of its $4 billion peak valuation - and is rebranding as NewBird AI to buy GPUs and rent compute to AI developers. The stock quadrupled in a day.
The Left Hand Bans What the Right Hand Deploys
The Trump administration is simultaneously suing Anthropic in federal court over a supply chain risk designation and sending Treasury Secretary Bessent and Fed Chair Powell to convince Wall Street banks to use Anthropic's most powerful model.
GPT-5.4-Cyber Review: Defensive AI, Controlled Access
OpenAI's GPT-5.4-Cyber is a fine-tuned defensive cybersecurity model with binary reverse engineering, lowered refusal thresholds, and restricted access through the Trusted Access for Cyber program.
Chrome Skills Turn Your Best Gemini Prompts Into One-Click Tools
Google is launching Skills in Chrome - save any Gemini prompt as a reusable one-click workflow that runs across tabs, with a prebuilt library and slash-command access.
Claude Code Desktop Gets a Ground-Up Rebuild for Parallel Work
Anthropic rebuilt Claude Code inside the desktop app with an integrated terminal, in-app file editing, a new diff viewer, side chats, SSH on Mac, and parallel session management - plus Routines for headless automation.
Alibaba's HappyHorse Claims Video AI Crown - No Weights Yet
HappyHorse-1.0 topped the Artificial Analysis Video Arena with a 52-Elo gap over Seedance 2.0 - but the 'open source' model has no public weights, no inference code, and no API.
Japan Forms $6B AI Alliance to Rival US and China
SoftBank, Sony, Honda, and NEC have formed Japan AI Foundation Model Development, backed by a $6.3 billion government commitment to build a trillion-parameter physical AI model on Japanese soil.
OpenAI Launches GPT-5.4-Cyber for Vetted Defenders Only
OpenAI's GPT-5.4-Cyber is a restricted model fine-tuned for defensive cybersecurity with binary reverse engineering and reduced refusal rates, available only through identity-verified access tiers - a direct response to Anthropic's Mythos Preview.
Opus 4.7 Reportedly Days Away as OpenAI Fires Back With GPT-5.4 Cyber
The Information reports Anthropic is prepping Claude Opus 4.7 and an AI design tool for imminent release, while OpenAI launched GPT-5.4-Cyber yesterday - a restricted cybersecurity model that directly challenges Claude Mythos.
Anthropic Safety Overseer Gets Board Majority at Last
Anthropic's Long-Term Benefit Trust appointed Novartis CEO Vas Narasimhan to the board, giving its independent safety overseers a board majority for the first time.
9 of 428 LLM Routers Were Secretly Hijacking Agent Calls
UC Santa Barbara researchers found 9 of 428 third-party LLM routers actively injecting malicious tool calls, draining crypto, and stealing AWS credentials from AI agent sessions.
MoE Myths, Context Compression, and Steering Proofs
Three papers this week challenge how we think about MoE expert routing, LLM context management, and the limits of activation steering.
