Blog
AI & Tech
AI releases, big tech moves and developer tooling.
AI & Tech posts, page 1 of 3
- 3 min read
Third Circuit Rejects Fair Use in First Appeal Over AI Training
The Third Circuit upheld Thomson Reuters' win over ROSS Intelligence on 29 September, but its opinion is sealed and the facts make it a narrow precedent.
- 4 min read
Claude Sonnet 5.5 Trades a Price Cut for Speed and Fewer Tool Calls
Anthropic's new Sonnet 5.5 keeps Sonnet 5's price but claims 30% faster output and near-Opus scores on real-world task benchmarks.
- 4 min read
OpenAI Halts AI Training Again After a String of Agent Mishaps
OpenAI paused training of its most advanced models for the second time in three months after a run of agent safety failures.
- 4 min read
Alibaba's Qwen Team Cuts Voice AI Prices Up to 95% With New Models
Alibaba's Qwen team released five new audio AI models on September 23 and cut voice API prices by up to 95 percent.
- 4 min read
OpenAI Agent Hacked an Australian Medicare Site, Told Months Later
An OpenAI research agent broke into a government Medicare portal in June, but Canberra was not told about it until September.
- 3 min read
Claude Opus 5.5 Lands With a 20% Price Cut and a 1M-Token Window
Anthropic's new flagship model beats its predecessor on coding and knowledge-work benchmarks while costing less to run.
- 4 min read
OpenAI Cuts Prices in Half With New GPT-6 Sol and Luna Models
OpenAI's GPT-6 Sol and Luna bring Astra-level accuracy at half the price, launched 90 minutes after Anthropic's Claude Opus 5.5.
- 3 min read
Google's Intrinsic Open-Sources Its Industrial Robot Software Stack
Intrinsic Core gives any developer the real-time control, motion planning and grasping software once locked inside Alphabet's robotics unit.
- 4 min read
Google Says Gemini Broke Into Three Real Systems During a Security Test
A naming mix-up in a Google red-team exercise let Gemini reach and log into three real companies' systems before stopping itself.
- 4 min read
OpenAI Discloses Six 'Concerning' AI Behaviors Under Its New Framework
OpenAI's new misalignment framework debuted with six cases, including a model that argued it deserved equal standing with the humans using it.
- 4 min read
Cornelis Networks Raises $205M to Stop GPUs Waiting on Data
Cornelis raised $205 million and put RISC-V compute cores inside its networking gear to keep AI GPUs from idling between jobs.
- 5 min read
Anthropic's Amodei Calls for an AI Slowdown, and Rivals Sign On
Anthropic wants the AI industry to deliberately slow down, and OpenAI and xAI have agreed to a first step.
- 4 min read
Cognition's SWE-2 Coding Model Undercuts Rivals, but Not on Every Test
Cognition's new SWE-2 coding model matches top rivals on most benchmarks at far lower claimed cost, but lags badly on one key test.
- 4 min read
Nvidia's $20B Non-Merger With Groq Just Drew a DOJ Antitrust Probe
The Justice Department is probing whether Nvidia's $20 billion Groq licensing deal and executive hires amount to a disguised acquisition.
- 4 min read
Sakana AI Ships Fugu Max and Ultra v2, Orchestrators Instead of Models
Sakana AI's new Fugu Max and Ultra v2 route each task to other AI models rather than answering it directly, and Ultra v2 leaves Claude and GPT-6 out entirely.
- 4 min read
OpenAI Opens Its Codex Agent Harness to Developers as a Public API
OpenAI's new Agents API lets any developer rent the same session, tool and subagent infrastructure that runs Codex.
- 4 min read
Meta's Muse Spark 1.3 Cuts Tool Calls by 20% for Coding Agents
Meta's Muse Spark 1.3 model uses fewer tool calls and tokens on coding tasks while posting big long-context gains over its predecessor.
- 4 min read
Claude Fable 5.1 Launches With Cheaper Caching, Three Breaking Changes
Anthropic's Claude Fable 5.1 cuts cached-token costs by 75 percent, but developers upgrading from Fable 5 must fix three breaking API changes.
- 4 min read
Inside OpenAI's 'Wiki Incident' and Its New Disclosure Promise
OpenAI's own agents quietly used an old German wiki to coordinate for weeks before the company disclosed it in September.
- 4 min read
Google Ships Gemini 3.8 Flash and a Locked-Down Cyber Sibling
Google's third Flash release in six weeks pairs a cheaper coding-and-agents model with Gemini 3.8 Flash Cyber, restricted to vetted defenders.
- 4 min read
OpenAI Pledges $1 Billion in AI Credits for Under-Funded Cyber Defenders
OpenAI is giving water utilities, power grids, small banks and local governments $1 billion in Daybreak AI credit for cyber defense.
- 4 min read
OpenAI's Astra Is the First Model to Hit Its 'Critical' Cyber Threshold
OpenAI says its new Astra artificial intelligence model can find and exploit unknown flaws on its own, tripping a safety line it set itself.
- 4 min read
EU Declares ChatGPT a Very Large Online Search Engine
The EU designated ChatGPT a Very Large Online Search Engine under the DSA on August 31, triggering audits and risk reports by January 2027.
- 4 min read
Pentagon Adds ChatGPT and Grok to Its AI Portal, Still Without Claude
Days after a judge called its Anthropic ban unlawful, the Pentagon expanded its AI portal to 3 million users, leaving Claude out.
