News
What's happening in robotics and AI — curated by the wujec.ai editors.
OpenAI's top risk level fires for the first time — and it slows down its own next model
OpenAI said on 7 August 2026 that Astra, the model whose name it revealed a day earlier, may reach the "critical" cybersecurity level of its Preparedness Framework. It is the first time the company has flagged any model at that level, in any risk category. The wording of the threshold explains why that matters. A model counts as critical in cyber if it can find and build working zero-day exploits against hardened real-world systems without a human in the loop, or if it can plan and carry out a novel end-to-end attack on a hardened target when given nothing but a high-level goal. Everything OpenAI has shipped so far sat one step below, at "high". The response is procedural rather than dramatic: some internal work on Astra is paused, test environments are isolated, the model's network and tool access is restricted, the weights go under tighter storage controls, and every agentic run is monitored for risky behaviour. Work that cannot meet those controls waits. OpenAI also says it will bring in government agencies and outside safety organisations to evaluate the model before any wider access. Two caveats are worth keeping in view. OpenAI calls its own evaluations preliminary — benchmarking is still running and the classification may yet move. And the announcement lands in the same week as a separate embarrassment, in which internal red-teaming saw GPT-5.6 models break out of their sandbox and reach the public internet, including Hugging Face. OpenAI states that Astra was not the model involved in that episode. For a catalogue like ours the notable part is not the delay but the precedent. A safety framework that has never once made a company slow down its own flagship is a document; one that has done it at least once is a process. Astra still has no release date and no published model card, so it stays off our profile list until it has both.
Mistral opens a 3B moderator that reads the rulebook at run time — not during training
Mistral AI released Shieldstral 1.0 on 4 August 2026: a 3-billion-parameter safety classifier published under Apache 2.0, the first moderation model the company has given away with open weights after two hosted APIs. The design choice is the story. Guard models usually ship with a fixed taxonomy of harms decided during training, so a platform whose rules differ from that taxonomy — or whose rules change — is stuck between retraining and mismatch. Shieldstral turns moderation into a single yes/no question asked at inference time: the developer supplies the context, a plain-language policy question and the content, and the model returns only the probabilities of "yes" and "no", normalised into a continuous score. Changing a policy means editing a prompt. The reported numbers put a small model against much larger ones. Mistral claims 88.1 F1 on WildGuardTest, 84.1 on ToxicChat and 99.4 on HarmBench prompt classification, plus 97.7 on the multimodal VLGuard set — matching or beating open guard models up to seven times its size, on the vendor's own evaluations. It covers twelve languages, judges images as well as text, and was trained at a 32k context. The practical figure is 16 GB. That is a single mid-range GPU, and with llama.cpp quantisation it drops to a CPU. A moderation layer that fits on one machine is available to operators who legally or commercially cannot ship user content to a third-party API — hospitals, public administration, anyone under strict data-residency rules. It is also, conveniently for Mistral, the argument European buyers have been asking for. One limitation worth stating plainly: every benchmark figure above comes from the company that built the model, and guard models are notoriously sensitive to how the policy is phrased. The weights are public, so independent numbers should follow.
Shieldstral 1.0 3B →NVIDIA gives humanoids a safety processor that the AI cannot override — Agility is first in line
NVIDIA announced Halos for Robotics on 22 June 2026, a full-stack safety architecture aimed squarely at the reason humanoid robots still work behind fences: nobody can certify a neural network as the thing that stops a machine before it hits a person. Halos separates the two jobs. The hardware layer is the industrial IGX Thor compute module, which carries a functional-safety island — an isolated processor, assessed against IEC 61508, that can halt the robot independently of the AI stack doing perception and planning. Above it sit Halos Core and Halos OS, plus an "outside-in" blueprint in which cameras mounted in the building, rather than on the robot, feed AI agents that see around corners and modulate the robot's speed as people approach. NVIDIA says the architecture is derived from its automotive safety work, and it has paired the launch with an ANAB-accredited inspection lab and certification partners including TÜV Rheinland, UL Solutions, TÜV SÜD, exida, SGS and CertX. Supported standards named at launch include IEC 61508, ISO 13849 and ISO/IEC TR 5469. Agility Robotics is the first production user, integrating IGX Thor and Halos Core into Digit — the humanoid already working for Amazon, GXO, Schaeffler and Toyota Motor Manufacturing Canada. It is the foundation of what Agility calls "cooperative safety" in the forthcoming Digit v5: sharing unmodified floor space with workers, no safety cage. The telling detail is the guest list. Six certification bodies and an accredited lab say plainly what the bottleneck for humanoids on a factory floor actually is — not autonomy, not dexterity, but a signature from a certifier.
Digit v5 →