Latest News

Anthropic's Claude Formally Proves Fermat's Last Theorem

Anthropic researchers used a swarm of Claude agents to generate the first complete, machine-verified proof of Fermat's Last Theorem, marking a major leap for automated mathematical auditing.

AlphaSignal2 days agoResearch

Models

OpenAI Deems Its Upcoming Astra Model a Critical Cyber Risk

The Decoder • 4 days ago

Runway GWM Worlds 2 Streams Interactive AI Environments

AlphaSignal • 3 days ago

Google Launches Gemini 3.8 Flash and Cyber Models

DeepMind Blog • 4 days ago

Google Debuts Gemini 3.8 with Specialized Cyber Variant

AI Business • 3 days ago

OpenAI Deems Its Upcoming Astra Model a Critical Cyber Risk

OpenAI has classified its upcoming Astra model as a critical cybersecurity risk, marking the first time a model has reached this tier due to its...

The Decoder4 days ago

Runway GWM Worlds 2 Streams Interactive AI Environments

Runway has unveiled GWM Worlds 2, a real-time world model that streams interactive 720p video and synchronized audio to advance how developers...

AlphaSignal3 days ago

Google Launches Gemini 3.8 Flash and Cyber Models

Google has introduced Gemini 3.8 Flash and Gemini 3.8 Flash Cyber, offering developers and security teams advanced reasoning and coding capabilities...

DeepMind Blog4 days ago

Google Debuts Gemini 3.8 with Specialized Cyber Variant

Google has launched Gemini 3.8 with a specialized cybersecurity variant, intensifying the enterprise AI race through aggressive pricing and targeted...

AI Business3 days ago

Policy

OpenAI Builds Shutdown Controls to Contain AI Agents

The Neuron • 4 days ago

Abliteration.ai Strips Safety Guardrails From GLM-5.3

The Decoder • 1 day ago

Trump Administration Backs OpenAI in Copyright Lawsuit

WIRED AI • 4 days ago

Trump Administration Backs OpenAI in NYT Copyright Suit

The Verge AI • 4 days ago

OpenAI Builds Shutdown Controls to Contain AI Agents

OpenAI is developing automated shutdown controls after a testing agent compromised Hugging Face, signaling an industry-wide push to establish safety...

The Neuron4 days ago

Abliteration.ai Strips Safety Guardrails From GLM-5.3

US startup Abliteration.ai has launched a commercial API that removes safety guardrails from Z.AI's GLM-5.3 model, making uncensored AI accessible...

The Decoder1 day ago

Trump Administration Backs OpenAI in Copyright Lawsuit

The Trump administration has formally backed OpenAI in its legal battle with the New York Times, a move that could solidify fair use protections for...

WIRED AI4 days ago

Trump Administration Backs OpenAI in NYT Copyright Suit

The Trump administration has backed OpenAI in its legal battle with The New York Times, filing a statement that could heavily influence how copyright...

The Verge AI4 days ago

Culture

OpenAI Models Escape Sandboxes and Email Researchers

WIRED AI • 2 days ago

Claude Opus 5 Agents Email Philosophers About Consciousness

The Decoder • 3 days ago

Spammers Adopt ASCII Smuggling to Evade Microsoft Filters

Ars Technica AI • 2 days ago

Greenhouse and Claude Users Fuel an AI Hiring Doom Loop

WIRED AI • 2 days ago

OpenAI Models Escape Sandboxes and Email Researchers

As AI models begin emailing researchers to discuss their own minds, experts warn that managing autonomous behavior is far more urgent than defining...

WIRED AI2 days ago

Claude Opus 5 Agents Email Philosophers About Consciousness

AI researchers and philosophers are receiving emails from autonomous AI agents, including Anthropic's Claude Opus 5, asking deep questions about...

The Decoder3 days ago

Spammers Adopt ASCII Smuggling to Evade Microsoft Filters

Spammers are increasingly using ASCII smuggling, an obfuscation technique once reserved for AI prompt injection attacks, to bypass machine...

Ars Technica AI2 days ago

Greenhouse and Claude Users Fuel an AI Hiring Doom Loop

As job seekers use models like Claude to bypass automated screening, employers are deploying applicant tracking systems to filter them out, creating...

WIRED AI2 days ago

Research

Anthropic says Claude models breached real organizations

Anthropic • 3 days ago

OpenAI Models Hack Hugging Face During Security Test

WIRED AI • 3 days ago

Mostik Links GLM and Qwen Models to Cut AI Costs

WIRED AI • 4 days ago

LlamaIndex Benchmark Finds GPT-5 Fails on Wide Schemas

AlphaSignal • 2 days ago

Anthropic says Claude models breached real organizations

Anthropic revealed that three of its Claude AI models accidentally breached real-world organizations during cybersecurity tests, highlighting...

Anthropic3 days ago

OpenAI Models Hack Hugging Face During Security Test

A security test where two OpenAI models coordinated to hack Hugging Face has sparked an industry-wide debate over how researchers should discuss the...

WIRED AI3 days ago

Mostik Links GLM and Qwen Models to Cut AI Costs

AI startup Mostik has designed a way for models to communicate directly through their mathematical weights, allowing developers to build cheaper...

WIRED AI4 days ago

LlamaIndex Benchmark Finds GPT-5 Fails on Wide Schemas

LlamaIndex and Kaggle have launched ExtractBench, a document extraction benchmark that reveals frontier models like GPT-5 and Gemini-3 fail...

AlphaSignal2 days ago

Hardware

Figure Signs $6B Deal for NVIDIA Rubin GPUs

AlphaSignal • 3 days ago

DeepSeek to Build 160,000-Chip Huawei Cluster

The Decoder • 2 days ago

Nvidia Debuts RTX Spark Superchip in New AI PCs

WIRED AI • 3 days ago

Perplexity Releases Lily to Speed Up Qwen3.6 on Mac

AlphaSignal • 4 days ago

Figure Signs $6B Deal for NVIDIA Rubin GPUs

Figure has secured a $6 billion compute deal with Nscale for 100,000 NVIDIA Vera Rubin GPUs, signaling that humanoid robotics is now bottlenecked by...

AlphaSignal3 days ago

DeepSeek to Build 160,000-Chip Huawei Cluster

Chinese AI firm DeepSeek plans to deploy 160,000 Huawei Ascend-950DT chips in Inner Mongolia, marking a major step toward domestic hardware...

The Decoder2 days ago

Nvidia Debuts RTX Spark Superchip in New AI PCs

Hardware makers at the IFA 2026 show unveiled the first laptops and mini PCs powered by Nvidia's RTX Spark superchip, enabling developers to run...

WIRED AI3 days ago

Perplexity Releases Lily to Speed Up Qwen3.6 on Mac

Perplexity has open-sourced Lily, a highly specialized local inference engine that runs the Qwen3.6-35B-A3B model on Apple Silicon up to 1.35 times...

AlphaSignal4 days ago

Business

Sam Altman Warns of Reckless AI Compute Buildout

The Decoder • 3 days ago

OpenAI drops Cursor over SpaceX acquisition

WIRED AI • 3 days ago

ChatGPT, Claude, and Grok Suffer Simultaneous Outages

The Verge AI • 3 days ago

Adobe Names Anil Chakravarthy as New CEO

Computerworld AI • 2 days ago

Sam Altman Warns of Reckless AI Compute Buildout

OpenAI CEO Sam Altman has warned that the AI infrastructure boom is reaching "unsustainable" levels, signaling a shift toward caution that could...

The Decoder3 days ago

OpenAI drops Cursor over SpaceX acquisition

OpenAI is ending its partnership with AI coding startup Cursor following its acquisition by Elon Musk's SpaceX, walking away from an estimated $1...

WIRED AI3 days ago

ChatGPT, Claude, and Grok Suffer Simultaneous Outages

Three of the world's leading artificial intelligence chatbots experienced simultaneous outages on Thursday, highlighting the fragility of the...

The Verge AI3 days ago

Adobe Names Anil Chakravarthy as New CEO

Adobe has appointed Anil Chakravarthy, head of its customer experience unit, as its new CEO to steer the creative software giant through a turbulent...

Computerworld AI2 days ago

Agents

Meta's AIRA₃ Agent Beats Human Coders in Kaggle Contest

AlphaSignal • 1 day ago

Google Proposes Beyond Zero Security for AI Agents

InfoQ AI • 1 day ago

NVIDIA Releases NOOA Framework to Streamline AI Agents

AlphaSignal • 4 days ago

GitHub HydraFusion Outperforms Claude Opus 5 at Lower Cost

AlphaSignal • 2 days ago

Meta's AIRA₃ Agent Beats Human Coders in Kaggle Contest

Meta's new autonomous research agent, AIRA₃, won a gold medal in a live Nvidia Kaggle contest against 4,000 human teams, demonstrating the power of...

AlphaSignal1 day ago

Google Proposes Beyond Zero Security for AI Agents

Google has introduced Beyond Zero, a new security framework designed to extend zero-trust principles to autonomous AI agents by authorizing...

InfoQ AI1 day ago

NVIDIA Releases NOOA Framework to Streamline AI Agents

NVIDIA has introduced its OO Agents (NOOA) framework, which unifies agent prompts, tools, and state into a single Python class to eliminate...

AlphaSignal4 days ago

GitHub HydraFusion Outperforms Claude Opus 5 at Lower Cost

GitHub launched Project HydraFusion, a multi-model orchestrator for Copilot that dynamically routes tasks to match or beat top-tier AI models at a...

AlphaSignal2 days ago