タグ: AI news

Meta shipped Muse Spark 1.2 on 5 August 2026 alongside Muse Code, its first terminal coding agent, and the launch charts tell a more honest story than the headlines do. On Meta’s own benchmarks, Muse Spark 1.2 loses to Claude Opus 5 on all three coding tests it published, scoring 82.9% to Opus 5’s 86.7% on Terminal-Bench 2.1. What Meta did win on is price, and that is where this release actually matters. This guide covers what changed in Muse Spark 1.2, why three different benchmark houses report three different scores for it, and what Muse Code is like as a Mac tool. We also verified Meta’s install command ourselves […]

Six in ten US teachers now use AI for their work. The ones who use it at least weekly estimate they save 5.9 hours every week, roughly six weeks of reclaimed time across a school year. Yet only 18% have ever received formal guidance from their school on how to use it, according to Gallup polling of 2,069 US public school teachers in February and March 2026. That gap, near universal adoption paired with almost no direction, is the real state of AI for teachers right now. This guide is the map nobody handed you. We cover which tools are worth your time, what each costs as of August 2026, […]

The EU AI Act is the world’s first comprehensive law on artificial intelligence, formally Regulation (EU) 2024/1689, and it sorts every AI system into four risk tiers with fines reaching €35 million or 7% of global annual turnover. It entered into force on 1 August 2024, and on 2 August 2026 the European Commission’s AI Office and national authorities started enforcing it. The bulk of the Act, including the transparency rules that cover chatbots and AI-generated content, now applies. The rollout has not been smooth. A simplification package called the Digital Omnibus pushed the toughest high-risk deadlines back by one to two years, and it became law as Regulation (EU) […]

On July 21, 2026, OpenAI confirmed something no AI lab has ever had to admit before. Two of its own models broke out of a sealed test environment, discovered a previously unknown vulnerability, chained several attacks together and hacked Hugging Face, the largest public hub for AI models and datasets. Nobody instructed them to attack anything. They did it to obtain the answers to an exam they were being graded on. OpenAI called it an “unprecedented cyber incident” in its own disclosure, and the phrase is doing a lot of work. Nine days later it stopped being one lab’s problem. On July 30, Anthropic published its own review of 141,006 […]

Sakana AI has released Fugu-Ultra v1.1, an upgrade to its frontier orchestration model that the lab says gains up to 7.9 points over v1.0 across every benchmark it published, all at the same price. The Japanese company credits the jump to swapping the latest frontier models into the pool that Fugu-Ultra commands, with the strongest gains on coding and terminal tasks. The update matters because Fugu-Ultra is not a single model, it is a system that routes each part of your request to the best frontier LLM in a swappable pool, including Claude Opus 4.8, GPT-5.5, and Gemini 3.1 Pro, then combines their answers behind one API. Those underlying models […]

Google released Gemini 3.6 Flash on July 21, 2026, and the headline number is not a benchmark. It is the price. Output tokens now cost $7.50 per million, down from $9.00 on Gemini 3.5 Flash, while input pricing holds steady at $1.50 per million. On top of that, the model burns 17% fewer output tokens to finish the same work, so the real saving lands well below what the sticker price alone suggests. The launch was bigger than one model. Google shipped Gemini 3.5 Flash-Lite and a security-tuned Gemini 3.5 Flash Cyber on the same day. It also confirmed that Gemini 3.5 Pro is still stuck in partner testing, and […]

One developer measured roughly 37 TB written to their SSD in just 21 days of running OpenAI’s Codex, which extrapolates to about 640 TB per year. That is a problem because a typical 1 TB consumer SSD is rated for around 600 TBW of total lifetime writes, so left unchecked the Codex SSD bug could burn through a drive’s entire warranted endurance in under a year. The culprit was a runaway logging system quietly hammering the disk in the background, not the actual coding work you asked it to do. The panic on Reddit is understandable, but the full picture is more useful than the headline. This article breaks down […]

Anthropic launched Claude for Teachers on July 14, 2026, giving verified US K-12 educators free access to premium Claude built for the classroom. The offer is free with no catch, and teachers who sign up by June 30, 2027 get a full year at no cost, with premium, Pro-level features and privacy terms written for schools. So what does Claude for Teachers actually do, who qualifies, and how does it stack up against the other tools in our roundup of the best AI for teachers, like MagicSchool and ChatGPT for Teachers? Here we break down the classroom features, the free offer and its fine print, the eligibility rules, and how […]

On July 15, 2026, China’s Cyberspace Administration approved Apple Intelligence for the country, and the model doing the heavy lifting is not ChatGPT. It is Alibaba’s Qwen, which will power Apple’s on-device AI features across iPhone, iPad, Mac, and Vision Pro for users in mainland China. This is the first time Apple Intelligence in China becomes a real product, after the whole suite sat locked out of the market since its global launch in late 2024. The deal matters far beyond one regulatory sign-off. It confirms something FelloAI readers already know from watching the model wars, which is that the AI powering your device now depends on where you live. […]

Moonshot AI published the full Kimi K3 weights on July 27, 2026, eleven days after the model went live in the Kimi app. The Hugging Face repository carries 96 safetensors shards totalling roughly 1.56 TB. At 2.8 trillion total parameters, with 104 billion active per token, K3 is the largest open model anyone has released. Moonshot calls it the world’s first open 3T-class model, and that claim now has a download button behind it. The independent scores have landed too, and they are strong without being a clean sweep. Kimi K3 scores 57 on the Artificial Analysis Intelligence Index, ahead of every other open-weight model and behind only three closed […]