Quirin Schlegel

Loading

The device picks the regime - Anthropic's hardware standard, and the deadline that is not 2028

Anthropic previewed the Model Hardware Standard on 27 August 2026 - a driver specification that lets an agent operate microscopes, liquid handlers, robotic arms and lasers through read and write primitives, standard device discovery and safety limits declared in the driver and enforced at the device, with QuEra's optical lock recovery moving from 58 per cent at about 150 seconds to 99.3 per cent across 700 trials. The obvious reading, that this is MCP for robots, misses the classification: Article 6(1) makes an agent enforcing a device's safety limits a safety component of an Annex I product, and the Digital Omnibus then moved Regulation (EU) 2023/1230 from Section A to Section B of Annex I, so that under Article 2(2) only Article 6(1), Article 60a and Articles 102 to 112 apply and the high-risk requirements arrive inside the Machinery Regulation by delegated act by 2 August 2028. The consequence is a date correction: the Machinery Regulation itself applies from 20 January 2027, eighteen months earlier, and its Annex I reaches safety components with fully or partially self-evolving behaviour using machine learning approaches ensuring safety functions, which is a notified body rather than a self-declaration - so the practitioner moves are to classify the device before wiring the agent, log who set each safety limit and what enforced it, ask the hardware vendor in writing for its January 2027 conformity route, and keep the driver's natural-language capability file as the intended-purpose document it already is.

1 September 202610 verified claims10 sources

Anthropic wired Claude to a laser last week. The compliance question that opens is not the one everyone is asking.

On 27 August, Anthropic previewed the Model Hardware Standard — a driver spec that lets an agent operate physical instruments: microscopes, liquid handlers, robotic arms, lasers. Read and write primitives, device discovery, safety limits declared in the driver and enforced at the device.

The numbers are Anthropic's own. At QuEra Computing, recovering an optical lock started at 58% and about 150 seconds. It ended at 99.3% across 700 trials — 695 recoveries. Over a 19-hour run, Claude's PID controllers never lost the lock; the expert-tuned ones dropped it about 1.6 times an hour.

Three things before you file this under "MCP for robots".

1. It makes the agent a safety component. AI Act Article 6(1): high-risk where the system "is intended to be used as a safety component of a product" covered by Annex I, and that product needs third-party conformity assessment. An agent enforcing a device's safety limits is that, by description.

2. Then the Act steps back. The Digital Omnibus moved Regulation (EU) 2023/1230 — machinery — from Section A to Section B of Annex I. For Section B, Article 2(2) leaves only Article 6(1), Article 60a and Articles 102 to 112 applying. The high-risk requirements arrive inside the Machinery Regulation instead, by delegated act, by 2 August 2028.

3. Which means your calendar has the wrong date. AI Act Annex I high-risk: 2 August 2028. The Machinery Regulation itself: 20 January 2027. Eighteen months earlier — and its Annex I reaches "safety components with fully or partially self-evolving behaviour using machine learning approaches ensuring safety functions", which is a notified body, not a self-declaration.

Monday:

1. Classify the device before you wire an agent to it. A robotic arm is machinery, Section B. An IVD analyser is Regulation 2017/746, Section A, and the whole high-risk chapter comes with it. 2. Log who set each safety limit and what actually enforced it — driver, firmware or model. 3. Ask your hardware vendor in writing which conformity route it plans for January 2027. 4. Keep the natural-language capability file the driver generates. It is your intended-purpose document whether you meant it or not.

The standard made the wiring cheap. It did not move the responsibility.

Which of your devices decides your regime?

#EUAIAct #AgenticAI #MachineryRegulation #AICompliance #Robotics

Evidence

What this rests on

Every factual claim in the edition above, stated on its own and checked before publication. If one of these is wrong, the argument is wrong — so they are listed where you can check them.

Anthropic previewed the Model Hardware Standard on 27 August 2026: a driver specification with read and write primitives, standard device discovery, and device characteristics and safety limits declared in an auto-generated reference file, reachable over MCP, a command line interface or code APIs.
Why it mattersIt is the moment agentic tooling stops being a software-supply-chain question and becomes a product-safety one. The driver, not the model, is what puts an agent inside a machine's safety envelope.
anthropic2026previewin
At QuEra Computing, re-acquisition of an optical lock started at 58% success and about 150 seconds and ended at 99.3% across 700 trials - 695 recoveries. Over a 19-hour run Claude's tuned PID controllers did not lose the lock once, against about 1.6 unlocks an hour for the expert-tuned controllers.
Why it mattersIt is the number that makes the wiring worth doing, which is what turns a research preview into deployed machinery inside the EU.
anthropic2026previewin
AI Act Article 6(1) classifies an AI system as high-risk where it is intended to be used as a safety component of a product, or is itself a product, covered by Union harmonisation legislation listed in Annex I, and that product is required to undergo third-party conformity assessment.
Why it mattersAn agent that enforces a device's declared safety limits meets condition (a) by description, without anyone deciding to build a high-risk system.
euaiactexpl2026classific
AI Act Article 2(2) provides that for high-risk AI systems classified under Article 6(1) relating to products covered by Section B of Annex I, only Article 6(1), Article 60a and Articles 102 to 112 apply.
Why it mattersIt means the AI Act's own high-risk chapter does not attach directly to AI in machinery - a practitioner reading Chapter III and planning against it is planning against the wrong instrument.
euaiactexpl2026scope
In AI Act Annex I, Regulation (EU) 2023/1230 on machinery is item 21 of Section B, while Regulation (EU) 2017/746 on in vitro diagnostic medical devices is item 12 of Section A.
Why it mattersSame driver, same agent, two different regimes: the device decides which one you are in, so classification has to happen before integration.
euaiactexpl2026annex
The Digital Omnibus moved Regulation (EU) 2023/1230 from Section A to Section B of Annex I to Regulation (EU) 2024/1689, on the reasoning that it is appropriate to move to a sectoral approach; the high-risk requirements are instead to be integrated into the Machinery Regulation by delegated act by 2 August 2028.
Why it mattersThe obligations do not disappear, they relocate - and they arrive through a machinery notified body rather than an AI Act conformity route.
eurogip2026machinery
Regulation (EU) 2023/1230 on machinery has been in force since 19 July 2023 and applies in general from 20 January 2027, replacing Directive 2006/42/EC.
Why it mattersIt is eighteen months before the AI Act's Annex I date of 2 August 2028, which is the date most practitioners have diarised.
eurlex2023regulatio · eurogip2026machinery
Annex I of Regulation (EU) 2023/1230 covers 'safety components with fully or partially self-evolving behaviour using machine learning approaches ensuring safety functions', and machinery embedding such systems; Annex I categories take the Article 25(2) and (3) conformity routes, which involve a notified body rather than internal production control.
Why it mattersIt is the operative sentence for anyone wiring a learning agent to a machine sold in the EU: self-declaration is not available.
bakermckenz2026machinery · nemkodigita2026guide · eurlex2023regulatio
The Annex I product route to high-risk classification under AI Act Article 6(1) applies from 2 August 2028 under Article 113 as amended.
Why it mattersIt is the date in most compliance calendars, and for machinery it is the later of the two that matter.
euaiactexpl2026classific
Anthropic states that MHS 'doesn't yet work with hardware that lacks a programming interface' and that Claude's spatial and physical reasoning 'have limitations that still require expert oversight'.
Why it mattersThe vendor is describing, in its own words, where Article 14 human oversight has to sit - which is a better starting point for an oversight measure than a generic policy.
anthropic2026previewin

Sources

AI Agent Storeaggregator
aiagentstor2026ai
Original Archived PDFretrieved 1 September 2026 · published 31 August 2026
Anthropicprimary source
anthropic2026previewin
Original Archived PDFretrieved 1 September 2026 · published 27 August 2026
Baker McKenziesecondary
bakermckenz2026machinery
Original Archived PDFretrieved 1 September 2026 · published 1 January 2026
EU AI Act Explorersecondary
Annex III: High-Risk AI Systems Referred to in Article 6(2)
Original Archived PDFretrieved 5 October 2026
EU AI Act Explorersecondary
Classification Rules for High-Risk AI Systems
Original Archived PDFretrieved 5 October 2026
EU AI Act Explorersecondary
euaiactexpl2026scope
Original Archived PDFretrieved 1 September 2026 · published 1 January 2026
EUR-Lexprimary source
eurlex2023regulatio
Original Archived PDFretrieved 1 September 2026 · published 14 June 2023
Eurogipsecondary
eurogip2026machinery
Original Archived PDFretrieved 1 September 2026 · published 6 August 2026
MarkTechPostaggregator
marktechpos2026anthropic
Original Archived PDFretrieved 1 September 2026 · published 29 August 2026
Nemko Digitalsecondary
nemkodigita2026guide
Original Archived PDFretrieved 1 September 2026 · published 1 January 2026

Every source for this edition, as one PDF BibTeX — all editions

Corrections

What changed after publication

No corrections. Three checks fired during research and are recorded here because each changed the edition: - The aggregator sweep dated the Model Hardware Standard to 31 August 2026; Anthropic's own post is dated 27 August 2026. The event date is used. - The same sweep listed a NIST agentic-identity item as 31 August news; the underlying NIST concept paper is from March 2026. Dropped. - QuEra's figures span stages: 96 per cent at about six seconds is an intermediate result, 99.3 per cent across 700 trials the endpoint. The post compares the start (58 per cent, about 150 seconds) with the endpoint and says so.