Who gives AI permission to act in the real world?
The flagship Atlas dossier on AI permission to act: identity, authorization, payments, context, trust, machine agency, human consent, delegation and responsibility when physical execution fails.
Maturity // UNVERIFIED
01 / The Atlas framing
Intelligence is not authority.
A system may understand exactly what should happen and still lack permission to make it happen.
This dossier tracks the transition from AI recommending an action to AI being authorized to execute it.
02 / Research areas under observation
- · Identity
- · Authorization
- · Payments
- · Location and context
- · Trust
- · Machine agency
- · Human consent
- · Service providers
- · Robots and autonomous systems
- · Verification
03 / Where the question actually lives
AI recommends an action
AI is authorized to execute it
Capability
Authority
04 / The physical scenario
- Can the agent spend €200?
- Can it order something to my home?
- Can it authorize transportation?
- Can it open a physical space?
- Can it delegate to another human?
- Can it delegate to another machine?
- Can it change a reservation?
- Can it initiate a physical service?
- Who is responsible when execution fails?
Research
Evidence // attributable claims only
This dossier is open and currently holds no research findings. Nothing is stated here as evidence until a primary source has been read and can be cited in full.
Source first. Claim second. Interpretation third.
- LEVEL 01 — Peer-reviewed research, arXiv from credible groups, university labs, conference proceedings
- LEVEL 02 — Official technical research from serious robotics / AI organisations
- LEVEL 03 — Technical interviews with researchers, engineers and founders
- LEVEL 04 — High-quality analysis from respected research institutions or investors
Required per source // source · institution · date · url · research type
Our read
Interpretation // thesis under investigation, not a finding
The future of AIis not only aboutwhat machines can do.It is aboutwhat we allow them to do.
05 / Related open problems
IDENTITY // unsolved
Who is an agent, legally and technically, when it acts for someone?
PAYMENTS // unsolved
Who gives an autonomous agent permission to spend money?
PERMISSIONS // unsolved
What are the limits of delegated authority for a physical action?
TRUST // unsolved
How do physical agents establish trust with people and businesses?
06 / Related Atlas nodes
IDENTITY → PERMISSIONS → PAYMENTS → ORCHESTRATION → VERIFICATION
- IDENTITY — Who an agent is when it acts on someone else's behalf in the world.
- PERMISSIONS — What an autonomous system is allowed to do, and who authorised it.
- PAYMENTS — How a machine pays for a real-world service under a mandate and a limit.
- ORCHESTRATION — Coordinating humans, providers, APIs and machines toward one outcome.
- VERIFICATION — Proving the intended physical outcome actually occurred.
07 / Related research
Atlas // research chain
- VERIFICATION
The dossiers are not isolated articles. Each one inherits the failure of the stage before it.
// Milan, Italy
// Real World Atlas
// Observation continues