Quirin Schlegel

Loading

Your agent just got a legal job description

China now regulates what an agent may decide alone; the EU deferred its high-risk deadlines and still has no agent-specific framework — so build the authority matrix yourself and be compliant in both.

27 July 20266 verified claims4 sources

Your AI agent just got a legal job description. China wrote it first.

On 15 July, the Cyberspace Administration's Implementation Opinions took effect — the first binding framework anywhere that regulates the *agent*, not the model. It splits agent decisions into three tiers of authority:

→ Tier 1 — human only. Signing contracts, moving funds, anything irreversible. → Tier 2 — the agent proposes, the user approves before execution. → Tier 3 — autonomous, but strictly inside an explicitly delegated scope.

The clause engineers keep skipping is Article 6: the boundaries of an agent's authority must be documented BEFORE deployment. Not reconstructed from logs after an incident. And liability follows whoever operated the agent — not whoever trained it.

Meanwhile, Brussels moved the goalposts. Parliament passed the Digital Omnibus on 16 June (423-57, 174 abstentions), the Council signed off on 29 June, and stand-alone high-risk obligations slid to 2 December 2027 — safety components to 2 August 2028. Transparency duties still land on 2 August 2026. And the AI Act still has no agent-specific framework at all.

So two of the three largest AI markets now disagree on the most basic question in agentic engineering: what is an agent allowed to decide by itself?

What I'd build on Monday, regardless of which regulator reaches you first:

1. Enumerate every irreversible action your agent can physically reach — payments, writes to prod, outbound comms, contract execution. 2. Put a typed approval gate in front of each one. Not a config flag. A type the compiler enforces. 3. Log the delegated scope, not just the tool call. "Agent called transfer()" is not evidence. "Agent called transfer() within scope X, delegated by user Y at time Z" is.

That artefact satisfies Beijing today and Brussels in eighteen months.

Deferral is not relief. Harmonised standards land close to the deadline — teams that wait will get weeks to adapt, not months. Building for agent authority now is the cheapest compliance work you'll ever do.

What tier is your agent operating at right now?

#AgenticAI #EUAIAct #AIGovernance #AIEngineering #Compliance

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.

China's Implementation Opinions on AI agents took effect on 15 July 2026, issued jointly by the CAC, the National Development and Reform Commission, and the Ministry of Industry and Information Technology.
Why it mattersSets the date anchor for the whole Agent Authority arc. Any post claiming "first binding agent framework" rests on this.
SRC - Pebblous - China AI agent decision tiers
Article 6 defines three tiers of agent decision authority: human-only for irreversible acts; agent-proposes-user-approves; and autonomous execution within an explicitly delegated scope, with the boundary documented before deployment.
Why it mattersThe tiering is the reusable engineering frame — it maps one-to-one onto Approval Gate and Delegated Scope.
SRC - Pebblous - China AI agent decision tiers
The liability principle attaches responsibility to whoever operated the agent, not whoever developed it.
Why it mattersInverts the usual model-vendor-blame framing and is directly relevant to integrators and freelancers who deploy other people's models.
SRC - Pebblous - China AI agent decision tiers
The Digital Omnibus deferred EU AI Act obligations for stand-alone high-risk AI systems to 2 December 2027.
Why it mattersThe headline number in every EU compliance conversation for the next eighteen months.
lathamandwa2026ai
Obligations for high-risk AI in regulated products and safety components were deferred to 2 August 2028.
Why it mattersRelevant to anyone embedding agents in regulated products — including building-sector software.
lathamandwa2026ai
AI Act transparency obligations still apply from 2 August 2026 and were not deferred — covering interactive systems, emotion recognition, and biometric categorisation.
Why it mattersThis is the deadline that actually binds this year, and the one most teams have stopped tracking because they heard 'deferred'.
lathamandwa2026ai

Sources

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

Corrections

What changed after publication

- **2026-07-27** — the draft asserted the Omnibus deferral as settled law. The Parliament's own legislative-train record showed the file only "close to adoption" as of 20 June. Verified via Freshfields that the Council gave final approval on **29 June 2026**. The post was rewritten to name the procedural chain (Parliament 16 June, 423-57-174; Council 29 June) rather than assert an outcome — more precise, and better positioning. Do not repeat the original phrasing.