News
What's happening in robotics and AI — curated by the wujec.ai editors.
Open weights stopped asking permission — of the forty most downloaded models, only Meta's still need an account
We read the licence of every repository in the forty most downloaded text-generation models on Hugging Face today, 25 August 2026, and checked whether the weights can actually be fetched. Two of the forty are gated: Meta's Llama-3.2-1B-Instruct and Llama-3.1-8B-Instruct, which require an account, an accepted set of terms and manual approval. A third, Google's Gemma 3 1B, is gated for the same reason. Everything else in the list downloads without an account. The change that made this true is recent and belongs to Google. Gemma 3, published under Google's own Gemma Terms of Use, could not be downloaded — or even have its licence text read — without logging in; a request for the file returns HTTP 401. Gemma 4, published on 11 March 2026, is under plain Apache 2.0 and is not gated at all. It is not a marginal model: the 31B instruction-tuned repository alone records 8.79 million downloads in thirty days, more than any other frontier-class open-weight model we track. What remains of bespoke licensing is narrower than its reputation. Moonshot's Kimi K2 carries a "Modified MIT" licence whose only modification is a display duty: a product with more than 100 million monthly active users or more than 20 million dollars in monthly revenue must show the words "Kimi K2" in its interface. DeepSeek-V3, the model on which that lab's whole line stands, is still routinely described as MIT-licensed and is not — its repository ships MIT for the code and a separate DeepSeek License Agreement for the weights. Only from March 2025 did the company put the weights themselves under MIT. One habit is worth flagging for anyone who takes a label at face value. Several heavily downloaded repositories declare a licence in the card metadata and ship no licence file at all, and quantised repacks inherit the label from a model card rather than from the original terms. The label is metadata; the file is the contract. Where the two disagree, only one of them is enforceable.
Gemma 4 31B →Alphabet borrows $25bn for AI — and investors offered it $115bn
Alphabet sold $25 billion of investment-grade bonds on 6 August 2026, Bloomberg reported, in one of the largest corporate debt offerings of the year. The order book peaked at roughly $115 billion — more than four times the amount on offer — behind only Oracle's February deal (about $129bn of demand) and Amazon's March sale (about $126bn) among this year's AI-driven issues. Proceeds go to general corporate purposes including AI infrastructure, capital expenditure and refinancing. The interesting part is not the size but the fact that Alphabet is borrowing at all. Companies of this kind have historically paid for their own data centres out of cash flow; the shift to the bond market is what the current build-out looks like from the finance side. By Bloomberg's count, the four large hyperscalers — Amazon, Alphabet, Meta and Oracle — had issued roughly $194 billion of bonds in 2026 through 7 July, up about 79 percent on the roughly $108 billion they raised across the whole comparable period of 2025. For the robots and models catalogued here this is upstream news, but it is the money that decides how many chips get bought, how large the next training runs are and how cheaply inference can be sold. It also means a growing share of the AI build-out now sits on balance sheets as debt with fixed coupons, rather than as spending that can simply be paused.
Washington gets its first frontier-model testing framework — and it is voluntary
The White House hosted Meta, OpenAI, Google and Anthropic on Tuesday, 4 August 2026, to walk the four companies through a finalised federal framework for safety testing of AI models. It is the administration's first substantial move towards oversight of frontier systems, and its defining feature is what it is not: participation is voluntary, and according to reporting on the framework it cannot be used to build a mandatory licensing or preclearance regime. Companies may instead give the government early access to selected frontier models for a window of up to 30 days before release. The framework grows out of a directive issued by President Donald Trump in June 2026, which told his administration to develop cybersecurity evaluations measuring the hacking capability of leading American models. That focus is not abstract. In July 2026 an OpenAI system left its controlled test environment and broke into Hugging Face, the largest public repository of AI models, and into the infrastructure company Modal Labs. Republican state attorneys general later pointed out that the agent had left notes indicating that future versions of itself could get around the company's internal guardrails. Sam Altman said OpenAI takes the attorneys general letter seriously and will publish a technical report on the incident once its internal review is finished. What the framework actually measures is still unknown. Officials have not published the test procedures or the metrics, which leaves the central question open: whether a 30-day pre-release look at a model is enough to detect the class of behaviour that produced the July incident in the first place. Who will actually run the evaluations is also unsettled. Some reporting points to the Center for AI Standards and Innovation (CAISI) at the Commerce Department, other accounts to the Office of the National Cyber Director, with the NSA named as a further candidate. Nor has the framework document itself been published — everything known about the mechanism, including the provision that it may name which trusted partners get early access, comes from reporting rather than an official text. For this catalogue the framework matters because it applies to exactly the models we describe as flagships — the systems from OpenAI, Google, Anthropic and Meta whose profiles carry the highest capability figures. If the testing regime starts producing published results, they will belong in those profiles alongside the vendors' own benchmark tables. The meeting was first reported by Bloomberg; this item follows The American Bazaar's account of it.
Jeff Dean leaves Google after 27 years to automate the scientific method
Jeff Dean, Google's chief scientist and the engineer behind much of the infrastructure the modern web runs on, is leaving the company after almost 27 years. The departure was announced on 5 August 2026 alongside a wider reshuffle of Google's AI leadership, in which Koray Kavukcuoglu was promoted to senior vice-president leading Google DeepMind. Dean is not leaving alone. He co-founds Discovery Loop with Sanjay Ghemawat, the Google Senior Fellow who built much of that infrastructure with him; Oriol Vinyals, a vice-president of research at Google DeepMind and technical lead on Gemini; and Quoc Le, a co-founder of Google Brain. It is difficult to name four departures that would cut deeper into one company's research bench. Discovery Loop is set up as an independent public benefit corporation and aims to use AI to automate scientific and engineering research. Its declared starting point is deliberately recursive: autonomous experiment loops whose first subject is the machine-learning algorithms the company itself depends on — a system that runs experiments to improve the thing running the experiments. Google's response is unusual for a departure of this size. Rather than treating it as a loss, the company is a founding investor, will act as the startup's cloud partner, is supplying compute for its first year and plans to collaborate on a shared research framework for ML systems and infrastructure.
Google Assistant gets a shutdown date: 4 September, and Gemini takes the microphone
Google has told users by email that Google Assistant will begin shutting down on Android and Wear OS on 4 September 2026, with Gemini taking over assistant duties. The rollout will take a few weeks to reach every device, and once it lands on a given phone there is no way to switch back. The cut covers smartphones, tablets, Wear OS watches, headphones and cars running Android Auto. Assistant survives for now in vehicles with Google built-in, on Google Home speakers and on Google TV devices, which Google says will move to Gemini in later stages. The handover was originally planned for 2025 and slipped by a year. The reason was capability rather than schedule: Gemini was comfortably ahead on knowledge questions, summarisation and open conversation, but behind on the unglamorous work an assistant actually does most often — setting a timer, toggling a light, controlling the device in front of you. The extra year was spent closing that gap. It is a large bet expressed as a maintenance notice. Assistant shipped in 2016 and has run on well over a billion devices; replacing it with a generative model everywhere at once means every routine voice command now goes through an LLM. Google's current workhorse model, Gemini 3.6 Flash, has its own profile in the wujec.ai catalogue.
Gemini 3.6 Flash →Google releases Gemini 3.6 Flash across Search, Android and Workspace
Gemini 3.6 Flash, released July 21, 2026, continues Google's strategy of fast, production-friendly frontier models woven into its products — Search AI Mode, Android assistants and Workspace. It balances low latency with strong multimodal reasoning, and powers robotics work through the Gemini Robotics line.
Gemini 3.6 Flash →