Skip to main content

When AI Agents Hold the Keys: Threat Model and Execution-Finality Requirements for Autonomous High-Consequence Systems
draft-das-agentic-effectuation-boundary-00

Document Type Active Internet-Draft (individual)
Author Sangam Das
Last updated 2026-09-19
RFC stream (None)
Intended RFC status (None)
Formats
Additional resources GitHub Composite Execution-Finality Reference for AI Agents and Distributed Systems
GitHub Runnable Reference Implementation and Adversarial Test Harness
Stream Stream state (No stream defined)
Consensus boilerplate Unknown
RFC Editor Note (None)
IESG IESG state I-D Exists
Telechat date (None)
Responsible AD (None)
Send notices to (None)
draft-das-agentic-effectuation-boundary-00
Network Working Group                                             S. Das
Internet-Draft                                    Independent Researcher
Intended status: Informational                         19 September 2026
Expires: 23 March 2027

   When AI Agents Hold the Keys: Threat Model and Execution-Finality
          Requirements for Autonomous High-Consequence Systems
               draft-das-agentic-effectuation-boundary-00

Abstract

   AI agents are increasingly being delegated authority to invoke APIs,
   move money, modify enterprise state, operate infrastructure, and
   trigger physical actions.  Existing mechanisms for authentication,
   authorization, proof-of-possession, attestation, request
   preconditions, and authorization-context propagation are necessary
   building blocks, but they do not by themselves establish a universal
   invariant that the exact consequential act approved earlier is the
   exact act permitted to become effective now.

   This document defines a threat model for autonomous high-consequence
   agents and identifies an effectuation-boundary gap: a request can be
   correctly authenticated, correctly authorized, correctly attested,
   and still be unsafe to execute because parameters, policy state,
   external prerequisites, delegation state, lineage, or the execution
   path changed after the earlier decision.  The document describes an
   execution-finality architecture in which a proposed act remains non-
   effective until a protected enforcement point verifies exact-act
   binding, freshness, current policy state, anti-replay state, required
   provenance, and path completeness immediately before effectuation,
   with authority consumed in coordination with the consequential
   commit.

Status of This Memo

   This Internet-Draft is submitted in full conformance with the
   provisions of BCP 78 and BCP 79.

   Internet-Drafts are working documents of the Internet Engineering
   Task Force (IETF).  Note that other groups may also distribute
   working documents as Internet-Drafts.  The list of current Internet-
   Drafts is at https://datatracker.ietf.org/drafts/current/.

   Internet-Drafts are draft documents valid for a maximum of six months
   and may be updated, replaced, or obsoleted by other documents at any
   time.  It is inappropriate to use Internet-Drafts as reference
   material or to cite them other than as "work in progress."

Das                       Expires 23 March 2027                 [Page 1]
Internet-Draft        Agentic Effectuation Boundary       September 2026

   This Internet-Draft will expire on 23 March 2027.

Copyright Notice

   Copyright (c) 2026 IETF Trust and the persons identified as the
   document authors.  All rights reserved.

   This document is subject to BCP 78 and the IETF Trust's Legal
   Provisions Relating to IETF Documents (https://trustee.ietf.org/
   license-info) in effect on the date of publication of this document.
   Please review these documents carefully, as they describe your rights
   and restrictions with respect to this document.  Code Components
   extracted from this document must include Revised BSD License text as
   described in Section 4.e of the Trust Legal Provisions and are
   provided without warranty as described in the Revised BSD License.

Table of Contents

   1.  Introduction  . . . . . . . . . . . . . . . . . . . . . . . .   4
     1.1.  Problem Sequence  . . . . . . . . . . . . . . . . . . . .   4
     1.2.  Industry-Standard Terminology and Functional
           Equivalence . . . . . . . . . . . . . . . . . . . . . . .   5
   2.  Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . .   6
   3.  Threat Model  . . . . . . . . . . . . . . . . . . . . . . . .   7
     3.1.  Assets  . . . . . . . . . . . . . . . . . . . . . . . . .   7
     3.2.  Security Boundary . . . . . . . . . . . . . . . . . . . .   7
   4.  High-Attention Threat Scenarios . . . . . . . . . . . . . . .   7
     4.1.  Scenario 1: The Perfectly Authenticated Fraudulent
            Transfer . . . . . . . . . . . . . . . . . . . . . . . .   7
     4.2.  Scenario 2: The Correct Breaker Command at the Wrong
            State  . . . . . . . . . . . . . . . . . . . . . . . . .   8
     4.3.  Scenario 3: The Robot Executes Yesterday's Safe Plan  . .   8
     4.4.  Scenario 4: Narrow Delegation Becomes Broad Effect  . . .   9
     4.5.  Scenario 5: A Retry Becomes a Double Consequence  . . . .   9
     4.6.  Scenario 6: The Emergency Path Becomes the Attack Path  .   9
     4.7.  Scenario 7: Attested Platform, Unauthorized
            Consequence  . . . . . . . . . . . . . . . . . . . . . .   9
     4.8.  Scenario 8: Missing Lineage Is Treated as No
            Restriction  . . . . . . . . . . . . . . . . . . . . . .  10
     4.9.  Scenario 9: Split-Brain Commit  . . . . . . . . . . . . .  10
     4.10. Scenario 10: Tool Schema Drift Changes the Meaning of the
            Same Call  . . . . . . . . . . . . . . . . . . . . . . .  10
   5.  Why OAuth Plus Sender-Constrained Authorization Does Not Alone
           Solve Irreversible Effectuation . . . . . . . . . . . . .  11
     5.1.  Holder-of-Key Is Not Exact-Act Finality . . . . . . . . .  11
     5.2.  RAR Can Describe the Act; Description Is Not the Commit
           Invariant . . . . . . . . . . . . . . . . . . . . . . . .  12

Das                       Expires 23 March 2027                 [Page 2]
Internet-Draft        Agentic Effectuation Boundary       September 2026

     5.3.  Token Replay Protection Is Not Effect Replay
           Protection  . . . . . . . . . . . . . . . . . . . . . . .  12
     5.4.  Authorization-Time State Is Not Necessarily
           Effectuation-Time State . . . . . . . . . . . . . . . . .  12
     5.5.  OAuth-Sufficient Deployments  . . . . . . . . . . . . . .  13
   6.  Why This Internet-Draft Is Different  . . . . . . . . . . . .  13
     6.1.  Proposed Interoperability Surface . . . . . . . . . . . .  14
     6.2.  Protocol-Level Distinguishing Invariants  . . . . . . . .  14
   7.  IETF Working-Group Relevance and Possible Dispatch  . . . . .  15
     7.1.  Likely Standards Decomposition  . . . . . . . . . . . . .  16
     7.2.  Question for Dispatch . . . . . . . . . . . . . . . . . .  17
   8.  Why Existing Controls Do Not Fully Close the Effectuation
           Boundary  . . . . . . . . . . . . . . . . . . . . . . . .  17
   9.  Execution-Finality Architecture . . . . . . . . . . . . . . .  18
     9.1.  Canonical Candidate Act . . . . . . . . . . . . . . . . .  19
     9.2.  Act-Bound Execution Authority . . . . . . . . . . . . . .  20
     9.3.  Effectuation Predicate  . . . . . . . . . . . . . . . . .  20
     9.4.  Check-and-Effect Semantics  . . . . . . . . . . . . . . .  21
   10. Required Security Properties  . . . . . . . . . . . . . . . .  22
   11. Domain Profiles . . . . . . . . . . . . . . . . . . . . . . .  22
     11.1.  Corporate Treasury and Autonomous Payments . . . . . . .  22
     11.2.  Power and Industrial Control . . . . . . . . . . . . . .  23
     11.3.  Robotic Logistics  . . . . . . . . . . . . . . . . . . .  23
   12. Industrial Relevance and Deployment Context . . . . . . . . .  23
     12.1.  OpenAI: Long-Running Agents and Tool Execution . . . . .  23
     12.2.  Anthropic: Autonomous Tool Use . . . . . . . . . . . . .  24
     12.3.  Microsoft: Autonomous Enterprise Agents  . . . . . . . .  24
     12.4.  Google Cloud: Managed Agent Runtimes and Agent-to-Agent
            Systems  . . . . . . . . . . . . . . . . . . . . . . . .  24
     12.5.  AWS: AgentCore and Enterprise Tool Connectivity  . . . .  25
     12.6.  NVIDIA: Robotics and Physical Actuation  . . . . . . . .  25
     12.7.  Apple: On-Device Models, Tool Calling, and App
            Actions  . . . . . . . . . . . . . . . . . . . . . . . .  25
     12.8.  Cross-Vendor Interoperability Significance . . . . . . .  26
     12.9.  Why a Standards-Level Treatment Matters  . . . . . . . .  26
   13. Adversarial Test Matrix . . . . . . . . . . . . . . . . . . .  27
   14. Relationship to Current IETF Work . . . . . . . . . . . . . .  28
   15. Operational Considerations  . . . . . . . . . . . . . . . . .  28
   16. Privacy Considerations  . . . . . . . . . . . . . . . . . . .  28
   17. Security Considerations . . . . . . . . . . . . . . . . . . .  29
   18. Open Questions for the IETF Community . . . . . . . . . . . .  29
   19. IANA Considerations . . . . . . . . . . . . . . . . . . . . .  29
   20. Additional Public Technical Resources . . . . . . . . . . . .  30
   21. Normative and Informative References  . . . . . . . . . . . .  30
   Author's Address  . . . . . . . . . . . . . . . . . . . . . . . .  31

Das                       Expires 23 March 2027                 [Page 3]
Internet-Draft        Agentic Effectuation Boundary       September 2026

1.  Introduction

   The security model of software is changing as AI systems move from
   generating content to initiating consequential operations.  An agent
   may prepare a payment, change a cloud route, dispatch a warehouse
   robot, alter an industrial setpoint, or invoke another privileged
   agent without a human reviewing every individual operation.

   The core vulnerability considered here is not simply that an AI model
   might be wrong.  The vulnerability is that a machine-generated act
   can cross from computation into an externally meaningful consequence
   even when the authorization evidence was created for a different act,
   a different state, an earlier policy epoch, an earlier dependency
   state, or a different execution path.

   Existing industry mechanisms address important parts of this problem.
   OAuth authorization constrains access to protected resources; Rich
   Authorization Requests can carry structured authorization details;
   DPoP sender-constrains tokens; Transaction Tokens propagate identity
   and authorization context through service chains; RATS and EAT
   provide evidence and claims about platform state; HTTP conditional
   requests can prevent operations against a changed target
   representation.

   Those controls remain useful and are not replaced by this document.
   The architectural delta is narrower: the component that is able to
   make the consequential operation effective independently verifies
   that the exact act being committed is still authorized under the
   current relevant state, and couples successful verification with
   single-use consumption or equivalent anti-reuse state.

1.1.  Problem Sequence

   The problem can be summarized as:

Das                       Expires 23 March 2027                 [Page 4]
Internet-Draft        Agentic Effectuation Boundary       September 2026

     autonomous computation
             |
             v
     proposed consequential act
             |
             v
     authentication / authorization / attestation
             |
             |   time passes; state changes; delegation changes;
             |   parameters may be substituted; retries may occur;
             |   another path may bypass the earlier check
             v
     EFFECTUATION BOUNDARY
             |
             +----> payment settles
             +----> breaker changes state
             +----> robot moves
             +----> data leaves trust boundary
             +----> privileged API mutates state

     Security question:
     Is the exact act crossing this boundary authorized NOW,
     under the state and constraints that matter to the consequence?

1.2.  Industry-Standard Terminology and Functional Equivalence

   The terminology in this document is intended to describe functions,
   not require a particular product name or implementation.  The same
   architecture may appear under different terminology in financial,
   industrial, telecom, cloud, safety, distributed-systems, or hardware-
   security environments.

   +===============+=========================+=========================+
   | Term used     | Equivalent or related   | Underlying function     |
   | here          | industry terms          |                         |
   +===============+=========================+=========================+
   | Candidate Act | transaction intent,     | The exact operation     |
   |               | command, job, API       | proposed before it is   |
   |               | mutation, control       | allowed to become       |
   |               | action                  | effective.              |
   +---------------+-------------------------+-------------------------+
   | Non-Effective | staged state, prepared  | The operation exists    |
   | State         | state, uncommitted      | but has not yet         |
   |               | state, pending command  | produced its governed   |
   |               |                         | consequence.            |
   +---------------+-------------------------+-------------------------+
   | Execution     | bounded capability,     | Cryptographic or        |
   | Handle        | transaction             | protected authority     |

Das                       Expires 23 March 2027                 [Page 5]
Internet-Draft        Agentic Effectuation Boundary       September 2026

   |               | authorization           | bound to an exact act   |
   |               | artifact, one-shot      | and constraints.        |
   |               | permit                  |                         |
   +---------------+-------------------------+-------------------------+
   | Finality Sink | policy enforcement      | The component or        |
   |               | point, reference        | boundary whose          |
   |               | monitor, command gate,  | successful action first |
   |               | actuation gate, safety  | makes the governed      |
   |               | interlock, commit gate  | consequence effective.  |
   +---------------+-------------------------+-------------------------+
   | Policy Epoch  | generation number,      | A version marker used   |
   |               | fencing epoch,          | to reject authority     |
   |               | configuration version   | created under stale     |
   |               |                         | policy or delegation    |
   |               |                         | state.                  |
   +---------------+-------------------------+-------------------------+
   | Act Binding   | transaction binding,    | Cryptographic           |
   |               | parameter binding,      | association between     |
   |               | request-object binding  | authority and the exact |
   |               |                         | canonical operation.    |
   +---------------+-------------------------+-------------------------+
   | Atomic        | check-and-commit,       | Prevents the same       |
   | Consume       | compare-and-swap,       | authority from becoming |
   |               | transactional consume,  | the basis for multiple  |
   |               | idempotency commit      | governed effects.       |
   +---------------+-------------------------+-------------------------+
   | Path          | complete mediation,     | Ensures alternate paths |
   | Completeness  | anti-bypass,            | cannot produce the same |
   |               | chokepoint enforcement  | consequence without the |
   |               |                         | required gate.          |
   +---------------+-------------------------+-------------------------+

                      Table 1: Functional Equivalence

2.  Scope

   This document focuses on machine-generated acts with potentially
   irreversible, high-cost, safety-relevant, or externally binding
   effects.  Examples include financial settlement, critical
   infrastructure control, robotic actuation, privileged enterprise
   changes, protected data release, and autonomous inter-agent
   delegation.

   This document does not standardize bank settlement protocols, grid
   protection algorithms, robot motion planning, model alignment, or
   domain-specific safety certification.  It defines a cross-domain
   security boundary and the properties expected at that boundary.

Das                       Expires 23 March 2027                 [Page 6]
Internet-Draft        Agentic Effectuation Boundary       September 2026

3.  Threat Model

   The threat model assumes that one or more upstream components can be
   buggy, compromised, stale, adversarially influenced, or operating on
   incomplete information.  The model does not require the AI model
   itself to be malicious.

   Relevant adversaries and failure sources include prompt injection,
   malicious tool output, compromised service workloads, stolen
   credentials, stale authorization context, policy revocation races,
   parameter substitution, replay, retry storms, split-brain services,
   schema drift, compromised plugins, malicious insiders, incomplete
   provenance, and alternate execution paths.

3.1.  Assets

   *  financial balances and settlement authority;

   *  industrial and grid control state;

   *  robotic motion and safety envelopes;

   *  privileged enterprise resources;

   *  confidential data and network egress;

   *  delegation state and policy configuration;

   *  auditability of machine-generated consequences.

3.2.  Security Boundary

   The protected boundary is the point at which a candidate operation
   first becomes externally meaningful.  A high-assurance deployment
   treats code that proposes the act as less trusted than the component
   that verifies and releases the effect.

4.  High-Attention Threat Scenarios

4.1.  Scenario 1: The Perfectly Authenticated Fraudulent Transfer

   A treasury agent is authorized to initiate vendor payments.  It holds
   valid OAuth credentials and a proof-of-possession key.  An attacker
   influences the model through an invoice attachment and causes the
   beneficiary account to be substituted after a legitimate approval
   workflow has completed.

Das                       Expires 23 March 2027                 [Page 7]
Internet-Draft        Agentic Effectuation Boundary       September 2026

   Authentication can still succeed.  Sender-constrained access can
   still succeed.  The agent can still be the legitimate software
   presenting the token.  The missing property is whether settlement
   authority is cryptographically bound to the exact beneficiary,
   amount, currency, invoice basis, policy epoch, and single-use
   transaction instance that the sink is about to commit.

     approved:
       pay(vendor=A, account=X, amount=100000)

                attacker-controlled substitution
                            |
                            v

     executed:
       pay(vendor=A, account=Y, amount=100000)

     If the sink validates only "agent may pay vendors",
     the authorization is valid while the act is wrong.

4.2.  Scenario 2: The Correct Breaker Command at the Wrong State

   A grid-management agent computes a breaker operation using telemetry
   observed at time T1.  Before the command reaches the actuator,
   topology, load, protection status, or operator override state changes
   at T2.  The command remains syntactically valid and the controller
   identity remains authorized.

   The safety question is therefore not only who issued the command, but
   whether the exact command remains admissible under the current
   control epoch and the current prerequisites that the deployment
   chooses to make execution-critical.

4.3.  Scenario 3: The Robot Executes Yesterday's Safe Plan

   A warehouse agent generates a route and grasp sequence while an aisle
   is clear.  A worker, another robot, or a pallet subsequently changes
   the local environment.  A stale prepared command is later delivered
   by a retrying service.

   Model correctness at planning time is insufficient.  A protected
   actuation gate can require freshness, a current safety epoch, command
   digest equality, single-use authority, and any domain-specific
   interlock result required by the robot controller.

Das                       Expires 23 March 2027                 [Page 8]
Internet-Draft        Agentic Effectuation Boundary       September 2026

4.4.  Scenario 4: Narrow Delegation Becomes Broad Effect

   Agent A delegates a task to Agent B, which invokes Agent C.  Each hop
   propagates valid identity and authorization context, but the final
   command includes parameters not present when the original delegation
   was evaluated.  This can occur through tool schema expansion, default
   arguments, data enrichment, or a compromised intermediate service.

   A final sink that verifies only the caller or scope may accept a
   semantically broader act.  Exact-act binding makes the final
   parameter set an object of verification rather than an assumption
   about well-behaved intermediaries.

4.5.  Scenario 5: A Retry Becomes a Double Consequence

   The agent submits a payment or actuation request.  The network times
   out after the underlying effect occurs but before the agent receives
   confirmation.  The agent retries.  Generic request authentication
   does not necessarily distinguish a legitimate retry from a second
   authorized effect.

   The execution authority therefore needs an idempotency or single-
   consumption semantic coordinated with the effect boundary.  For
   physical systems where rollback is impossible, the protected state
   transition must prevent command re-release even if higher layers
   repeat the request.

4.6.  Scenario 6: The Emergency Path Becomes the Attack Path

   The primary API passes through a final enforcement gate, but a legacy
   interface, local maintenance command, direct message bus, or
   privileged recovery path can produce the same consequence without
   that gate.  An attacker routes the operation through the uncontrolled
   path.

   This is a complete-mediation problem.  Execution-finality is not
   established if a governed consequence has an alternate effectuation
   path that is outside the required enforcement set.

4.7.  Scenario 7: Attested Platform, Unauthorized Consequence

   A workload presents fresh attestation evidence showing approved
   software and expected platform state.  The workload then asks to
   perform an operation that is outside the intended transaction or uses
   stale delegation state.

Das                       Expires 23 March 2027                 [Page 9]
Internet-Draft        Agentic Effectuation Boundary       September 2026

   Attestation answers an important question about the entity and its
   state.  It does not by itself establish that a specific consequential
   act is authorized for effectuation.  The architecture can use RATS/
   EAT evidence as an input while still requiring act-specific
   authorization at the sink.

4.8.  Scenario 8: Missing Lineage Is Treated as No Restriction

   A downstream service receives an act without required provenance or
   delegation lineage.  If missing lineage is interpreted as an empty
   restriction set, stripping metadata can increase authority.

   A safer semantic is to distinguish EMPTY from UNKNOWN.  EMPTY means
   the verified lineage is intentionally empty.  UNKNOWN means the
   required lineage cannot be established.  For consequences whose
   policy requires lineage, UNKNOWN remains non-effective.

4.9.  Scenario 9: Split-Brain Commit

   One subsystem records that an operation failed while another
   subsystem has already committed the consequence.  A compensating
   retry or secondary controller then creates an inconsistent or
   duplicated state.

   Deployments should define which state transition is authoritative,
   how idempotency survives crashes, and whether the authority
   consumption record and the consequential commit can be made atomic or
   recoverably coupled.

4.10.  Scenario 10: Tool Schema Drift Changes the Meaning of the Same
       Call

   An agent was approved to invoke a tool when version 7 of the schema
   mapped a field to a read-only operation.  Version 8 changes a
   default, expands a wildcard, or maps the same high-level request to a
   state-changing operation.

   Authority can therefore be bound not only to textual arguments but
   also to the tool or mapping revision whose semantics were evaluated.
   Digest or version mismatch causes re-evaluation rather than dynamic
   assumptions of equivalence.

Das                       Expires 23 March 2027                [Page 10]
Internet-Draft        Agentic Effectuation Boundary       September 2026

5.  Why OAuth Plus Sender-Constrained Authorization Does Not Alone Solve
    Irreversible Effectuation

   OAuth is directly relevant to this problem, and this document does
   not claim that OAuth authorization is weak or inappropriate.  OAuth
   answers delegation and protected-resource access questions.  Rich
   Authorization Requests can express fine-grained transaction details.
   DPoP and mutual-TLS certificate-bound access tokens can make an
   access token sender-constrained, sometimes informally described as
   "non-bearer", so possession of a copied token alone is insufficient.

   The remaining problem appears when a valid authorized request is able
   to trigger an irreversible or externally binding effect.  Sender
   constraint proves that the presenter possesses the required key.  It
   does not, by itself, prove that the irreversible effect about to
   occur is still the exact effect whose semantics were evaluated, that
   all execution-critical dependencies remain current, or that the
   authority and effect will be consumed as one recoverable state
   transition.

5.1.  Holder-of-Key Is Not Exact-Act Finality

   A DPoP proof binds an OAuth presentation to a key and to selected
   HTTP request properties.  A certificate-bound token similarly binds
   token use to the holder of the corresponding private key.  These are
   strong defenses against stolen-token use.  However, the legitimate
   key holder can still be a buggy, compromised, stale, or adversarially
   influenced agent.

   The relevant distinction is:

     sender-constrained authorization asks:

         "Is this token being presented by the legitimate key holder,
          for an allowed protected-resource request?"

     execution-finality additionally asks:

         "Is this exact consequential act, with these final semantics,
          permitted to become effective NOW, under the current
          execution-critical state, and can this authority create
          no unintended second effect?"

Das                       Expires 23 March 2027                [Page 11]
Internet-Draft        Agentic Effectuation Boundary       September 2026

5.2.  RAR Can Describe the Act; Description Is Not the Commit Invariant

   RFC 9396 can carry highly specific authorization details, including
   payment amount and creditor information.  If an authorization profile
   includes every execution-relevant field, and the resource server
   verifies those exact fields at the final commit boundary, then simple
   parameter substitution can already be prevented.  This document
   explicitly relies on that capability where available.

   The residual case is state or semantics not fully represented by the
   authorization object: a beneficiary alias resolves differently, a
   tool or mapping revision changes, a delegation or policy epoch
   advances, an external prerequisite changes, a physical safety state
   changes, or an alternate path performs the same consequence without
   traversing the resource server check.

   Therefore, the delta is not "OAuth cannot describe a payment".  The
   delta is that authorization description alone does not define a
   cross-domain commit-time invariant covering current dependencies,
   irreversible effect, anti-replay state, and path completeness.

5.3.  Token Replay Protection Is Not Effect Replay Protection

   DPoP proof freshness, token sender constraint, and token replay
   defenses operate on protocol credentials and requests.  A different
   failure occurs when the first request commits the external effect but
   the response is lost.  A legitimate client may then construct a new
   valid proof and retry the authorized request.

   For an irreversible effect, the system needs an application or sink
   invariant such as a transaction identifier, idempotency key, one-shot
   capability, protected replay store, or equivalent state that is
   coordinated with the actual commit.  Preventing reuse of a stolen
   credential and preventing duplication of an already committed
   consequence are related but distinct properties.

5.4.  Authorization-Time State Is Not Necessarily Effectuation-Time
      State

   An authorization server can make a correct decision at time T1.  The
   resource server can receive a valid sender-constrained token at T2.
   Between T1 and the irreversible effect at T3, an external
   prerequisite, policy generation, delegation, device state, mapping,
   or safety condition can change.

Das                       Expires 23 March 2027                [Page 12]
Internet-Draft        Agentic Effectuation Boundary       September 2026

     T1  AS authorizes a specific operation
             |
             |    valid RAR / token / sender constraint
             v
     T2  RS accepts a request from the legitimate key holder
             |
             |    relevant external or local state changes
             v
     T3  irreversible effect becomes real
             |
             +--> settlement
             +--> actuator release
             +--> protected data egress
             +--> destructive enterprise mutation

     The property introduced here is a defined check at T3, not merely
     another credential evaluated at T1 or T2.

5.5.  OAuth-Sufficient Deployments

   There is an important limiting case.  If an OAuth deployment already
   binds every execution-relevant field, verifies those bindings at the
   component that performs the final effect, revalidates every required
   changing prerequisite, prevents alternate bypass paths, and couples
   idempotent or single-use authorization state to the actual commit,
   then that deployment already implements most of the execution-
   finality invariant described here.

   In that case, this document does not require replacing OAuth.  Its
   value is to make those properties explicit, portable, testable, and
   applicable to systems whose final consequence is not merely an HTTP
   resource response, including message-driven services, hardware
   control, robotics, industrial actuation, and autonomous settlement.

6.  Why This Internet-Draft Is Different

   This document is not a new general authorization framework.  It does
   not define who a principal is, how a user delegates access, how
   consent is collected, or how a model decides what action to propose.
   It starts after those mechanisms have produced an apparently
   authorized candidate act.

   The document attempts to standardize the security properties of the
   final transition from an authorized proposal to an effective
   consequence.  That makes the unit of analysis the effectuation
   boundary rather than the login session, access token, model output,
   API call, or attested workload.

Das                       Expires 23 March 2027                [Page 13]
Internet-Draft        Agentic Effectuation Boundary       September 2026

     EXISTING LAYERS                         THIS DOCUMENT

     identity  ---------------------+
     authentication ----------------+----+
     OAuth delegation --------------+    |
     RAR / scopes ------------------+    |
     sender constraint -------------+    |  inputs
     workload identity -------------+    |
     attestation -------------------+    |
     safety / policy state ---------+----+
                                         |
                                         v
                                +------------------+
                                | EFFECTUATION GATE|
                                | exact act?       |
                                | current state?   |
                                | current epoch?   |
                                | required lineage?|
                                | unused authority?|
                                | governed path?   |
                                +--------+---------+
                                         |
                                         v
                                    IRREVERSIBLE
                                     CONSEQUENCE

6.1.  Proposed Interoperability Surface

   The candidate interoperability surface is deliberately small: an
   exact-act commitment, an authority identifier or nonce, a sink
   identifier, freshness, relevant epoch or dependency commitments,
   constraints, and a verifiable result or receipt where required.
   Existing OAuth, COSE, RATS, or workload-identity artifacts can carry
   or supply these values if the responsible working groups determine
   that profiling existing mechanisms is preferable to defining new
   protocol objects.

6.2.  Protocol-Level Distinguishing Invariants

   1.  The candidate act is explicitly non-effective before final
       verification.

   2.  The final sink reconstructs or obtains the exact act that will
       become effective.

   3.  Authority is compared against that exact act, not merely against
       caller identity or broad scope.

Das                       Expires 23 March 2027                [Page 14]
Internet-Draft        Agentic Effectuation Boundary       September 2026

   4.  Selected mutable state is revalidated at the effectuation
       boundary.

   5.  Authority reuse and crash/retry behavior are defined relative to
       the actual consequence.

   6.  Missing required provenance is UNKNOWN and fails closed rather
       than silently widening authority.

   7.  All materially equivalent effectuation paths are within the
       stated enforcement coverage.

7.  IETF Working-Group Relevance and Possible Dispatch

   The problem is cross-area because an autonomous act can begin as an
   OAuth-authorized API operation, traverse several workloads, consume
   attestation evidence, cross a CBOR/COSE boundary, and terminate at a
   database, payment rail, gateway, or physical controller.  The
   document therefore separates the common invariant from any single
   encoding or transport.

    +==========+======================+===============================+
    | Group or | Relevant existing    | Potential relationship to     |
    | venue    | work                 | this draft                    |
    +==========+======================+===============================+
    | OAuth    | delegated            | Primary integration point for |
    |          | authorization, RAR,  | authorization inputs and      |
    |          | DPoP, token          | multi-hop authorization       |
    |          | exchange,            | context.  A profile could     |
    |          | Transaction Tokens   | bind OAuth authorization to   |
    |          |                      | an exact execution commitment |
    |          |                      | without redefining OAuth.     |
    +----------+----------------------+-------------------------------+
    | WIMSE    | workload identity    | Relevant when an agentic act  |
    |          | and least-privilege  | crosses several workloads and |
    |          | access across multi- | the final sink must           |
    |          | service environments | distinguish workload identity |
    |          |                      | from authority for the final  |
    |          |                      | consequence.                  |
    +----------+----------------------+-------------------------------+
    | RATS     | remote attestation   | Attestation can supply        |
    |          | evidence,            | trustworthy state predicates  |
    |          | attestation results, | to the finality sink.  This   |
    |          | appraisal, epoch-    | draft keeps attestation       |
    |          | related state        | evidence separate from act    |
    |          |                      | authorization.                |
    +----------+----------------------+-------------------------------+
    | COSE     | CBOR signing, MAC,   | Relevant if an execution      |

Das                       Expires 23 March 2027                [Page 15]
Internet-Draft        Agentic Effectuation Boundary       September 2026

    |          | encryption, keys,    | handle or finality receipt    |
    |          | and registered COSE  | needs a compact integrity-    |
    |          | attributes           | protected CBOR representation |
    |          |                      | or new registered attributes. |
    +----------+----------------------+-------------------------------+
    | ACE /    | authorization and    | Potential profile venue for   |
    | CoRE     | constrained          | constrained controllers,      |
    |          | application          | robots, gateways, or          |
    |          | protocols for        | industrial devices where the  |
    |          | resource-constrained | effectuation sink is not a    |
    |          | environments         | conventional web service.     |
    +----------+----------------------+-------------------------------+
    | DISPATCH | routing new cross-   | Suitable first venue if no    |
    | /        | cutting work and     | current WG owns the cross-    |
    | security | identifying overlap  | domain effectuation-boundary  |
    | dispatch | with existing WGs    | invariant.                    |
    | process  |                      |                               |
    +----------+----------------------+-------------------------------+
    | SAAG     | cross-cutting        | Useful for threat-model       |
    |          | Security Area        | review and architectural      |
    |          | discussion           | criticism.  SAAG is a         |
    |          |                      | discussion forum rather than  |
    |          |                      | a document-adopting WG.       |
    +----------+----------------------+-------------------------------+

                    Table 2: Relationship to IETF Groups

7.1.  Likely Standards Decomposition

   The work can be decomposed rather than forcing one working group to
   standardize every layer:

   1.  *Architecture / threat model:* define the effectuation boundary,
       attacker model, and invariants.

   2.  *Authorization profile:* define how existing OAuth authorization
       details or transaction context bind to an exact act when OAuth is
       used.

   3.  *Attestation binding:* define how RATS results or epoch evidence
       become execution predicates without turning attestation into
       authorization.

   4.  *Compact object format:* reuse COSE/CBOR only if a portable
       execution-handle or receipt representation is needed.

Das                       Expires 23 March 2027                [Page 16]
Internet-Draft        Agentic Effectuation Boundary       September 2026

   5.  *Constrained-device profile:* map the invariant to CoAP/ACE or
       device-local enforcement when an irreversible act occurs outside
       an HTTP service.

7.2.  Question for Dispatch

   The central dispatch question is therefore not whether the IETF
   should invent another authorization protocol.  It is whether existing
   IETF mechanisms, when combined, already define an interoperable and
   testable invariant for the final irreversible effect, and if not,
   which existing working group or new narrowly scoped effort should own
   that invariant.

8.  Why Existing Controls Do Not Fully Close the Effectuation Boundary

   The mechanisms below are complementary.  This section describes the
   remaining boundary condition rather than a defect in those protocols.

Das                       Expires 23 March 2027                [Page 17]
Internet-Draft        Agentic Effectuation Boundary       September 2026

    +===============+====================+============================+
    | Mechanism     | Primary property   | Residual question at       |
    |               |                    | effectuation               |
    +===============+====================+============================+
    | OAuth         | Delegated access   | Does current authority     |
    | authorization | to a protected     | bind the exact final act   |
    |               | resource.          | and current execution-     |
    |               |                    | critical state?            |
    +---------------+--------------------+----------------------------+
    | RAR           | Structured         | Are the committed          |
    |               | authorization      | parameters exactly those   |
    |               | details.           | details, and are later-    |
    |               |                    | changing prerequisites     |
    |               |                    | still satisfied?           |
    +---------------+--------------------+----------------------------+
    | DPoP          | Sender-constrained | Does the legitimate sender |
    |               | token              | possess authority for this |
    |               | presentation.      | exact consequential act?   |
    +---------------+--------------------+----------------------------+
    | Transaction   | Propagation of     | Is propagated context      |
    | Tokens        | identity and       | sufficient for a commit-   |
    |               | authorization      | time decision at the final |
    |               | context through a  | consequence boundary?      |
    |               | call chain.        |                            |
    +---------------+--------------------+----------------------------+
    | RATS / EAT    | Evidence and       | Is this exact act          |
    |               | claims about       | authorized, current,       |
    |               | entity/platform    | single-use, and path-      |
    |               | state.             | complete?                  |
    +---------------+--------------------+----------------------------+
    | HTTP If-Match | Conditional method | What about non-resource    |
    |               | execution against  | dependencies, multiple     |
    |               | a matching target  | resources, delegation      |
    |               | representation.    | epochs, physical           |
    |               |                    | actuation, or consequences |
    |               |                    | outside one HTTP origin?   |
    +---------------+--------------------+----------------------------+

       Table 3: Complementary Controls and Residual Boundary Question

9.  Execution-Finality Architecture

   The proposed architecture separates the ability to compute an act
   from the authority required to make that act effective.

Das                       Expires 23 March 2027                [Page 18]
Internet-Draft        Agentic Effectuation Boundary       September 2026

       +---------------------+
       | Model / Agent / App |
       +----------+----------+
                  |
                  | proposes exact Candidate Act
                  v
       +---------------------+
       | Non-Effective State |
       +----------+----------+
                  |
                  | protected validation
                  v
       +-----------------------------+
       | Authorization / Policy /    |
       | Attestation / Safety Inputs |
       +--------------+--------------+
                      |
                      | issue act-bound authority
                      v
          +-------------------------+
          | Execution Handle        |
          | digest + nonce + epoch  |
          | constraints + expiry    |
          +------------+------------+
                       |
                       v
          +-------------------------+
          | FINALITY SINK           |
          | - reconstruct act       |
          | - compare digest        |
          | - verify freshness      |
          | - verify current epoch  |
          | - verify required state |
          | - verify lineage        |
          | - consume once          |
          +------------+------------+
                       |
             success only
                       v
                EFFECTUATION

9.1.  Canonical Candidate Act

   A deployment defines a deterministic representation of the execution-
   relevant fields.  One abstract form is:

Das                       Expires 23 March 2027                [Page 19]
Internet-Draft        Agentic Effectuation Boundary       September 2026

     A = Canon(
           operation,
           target,
           arguments,
           principal,
           delegated_by,
           purpose,
           budget,
           destination,
           tool_or_schema_revision,
           policy_epoch,
           dependency_commitments,
           sink_id
         )

     act_digest = H(A)

   Canonicalization is application-specific and must avoid ambiguous
   encodings.  Fields that can change the consequence need to be
   represented either directly or by an unambiguous commitment.

9.2.  Act-Bound Execution Authority

   A representative execution handle can be modeled as:

     EH = Protect(
           act_digest,
           authority_id,
           nonce,
           issued_at,
           expires_at,
           policy_epoch,
           constraint_set,
           sink_id
         )

   Protect() can be implemented with an integrity-protected token,
   protected local object, hardware-bound capability, or another
   construction that provides the deployment's required authenticity and
   anti-forgery properties.

9.3.  Effectuation Predicate

   Let S_now be the protected state visible to the finality sink.  A
   candidate act A becomes eligible for effectuation only if the
   required predicates are true:

Das                       Expires 23 March 2027                [Page 20]
Internet-Draft        Agentic Effectuation Boundary       September 2026

     Permit(A, EH, S_now) =

         ValidProtection(EH)
      AND H(Canon(A)) == EH.act_digest
      AND Fresh(EH)
      AND NotConsumed(EH.nonce)
      AND EH.policy_epoch == S_now.policy_epoch
      AND PolicyAllows(A, S_now)
      AND RequiredDependenciesCurrent(A, S_now)
      AND RequiredLineage(A) != UNKNOWN
      AND SinkMatches(EH.sink_id)
      AND PathIsGoverned(A)

     Effectuate(A) only if Permit(...) == TRUE.

9.4.  Check-and-Effect Semantics

   function effectuate(candidate_act, execution_handle):
       act = canonicalize(candidate_act)

       begin protected_transition:
           assert verify_handle(execution_handle)
           assert hash(act) == execution_handle.act_digest
           assert current_time <= execution_handle.expires_at
           assert replay_store.is_unused(execution_handle.nonce)
           assert policy_epoch() == execution_handle.policy_epoch
           assert current_policy_allows(act)
           assert required_dependencies_are_current(act)
           assert required_lineage(act) != UNKNOWN
           assert sink_id() == execution_handle.sink_id

           reserve_or_mark_inflight(execution_handle.nonce)

           result = commit_governed_effect(act)

           if result == COMMITTED:
               replay_store.mark_consumed(execution_handle.nonce)
               protected_commit()
               return SUCCESS

           protected_abort_or_recover()
           return FAILURE

   Exact crash semantics depend on the underlying consequence.  For a
   database or payment rail, transactional or idempotent commit
   primitives may exist.  For irreversible physical actuation, the
   protected state machine should ensure that a crash cannot cause the
   same authority to release the command a second time.

Das                       Expires 23 March 2027                [Page 21]
Internet-Draft        Agentic Effectuation Boundary       September 2026

10.  Required Security Properties

   A system claiming an execution-finality property for a governed class
   of acts should make the following properties externally reviewable:

   1.  *EF-1 Exact-Act Binding:* authority is bound to a deterministic
       representation of the execution-relevant act.

   2.  *EF-2 Commit-Time Revalidation:* state designated as execution-
       critical is checked at or immediately before the effectuation
       boundary.

   3.  *EF-3 Anti-Replay:* the same authority cannot silently create
       multiple governed effects unless multiplicity is explicitly
       authorized.

   4.  *EF-4 Epoch/Fencing:* revocation, policy change, delegation
       change, or schema revision can invalidate previously issued
       authority when the deployment requires it.

   5.  *EF-5 Fail-Closed Unknowns:* required but unavailable lineage or
       state is represented as UNKNOWN rather than inferred to mean
       unrestricted authority.

   6.  *EF-6 Path Completeness:* all paths capable of producing the
       governed consequence are included in the enforcement model or
       explicitly identified as exceptions.

   7.  *EF-7 Sink-Local Verification:* the effectuation component does
       not rely solely on an upstream statement that validation
       occurred.

   8.  *EF-8 Recoverable Commit Semantics:* crash, timeout, retry, and
       split-brain behavior cannot silently turn one authorized act into
       multiple effects.

11.  Domain Profiles

11.1.  Corporate Treasury and Autonomous Payments

   A financial profile can bind beneficiary identifier, destination
   account, amount, currency, fee bound, invoice or business basis,
   source account, approval class, settlement rail, policy epoch,
   transaction nonce, and any risk or human-approval evidence required
   by the deployment.

Das                       Expires 23 March 2027                [Page 22]
Internet-Draft        Agentic Effectuation Boundary       September 2026

11.2.  Power and Industrial Control

   A control profile can bind device or zone, command type, setpoint,
   allowed range, control epoch, topology or configuration commitment,
   maintenance state, operator override state, and required domain-
   specific safety interlock result.  Execution-finality is not a
   substitute for protection relays or safety engineering; it is a gate
   that can make selected safety and authorization predicates
   technically necessary for command release.

11.3.  Robotic Logistics

   A robotics profile can bind robot identity, command sequence,
   workspace or zone, payload class, maximum speed or force envelope,
   route revision, local safety epoch, freshness window, and the
   actuation sink expected to consume the authority.

12.  Industrial Relevance and Deployment Context

   The effectuation-boundary problem is relevant to current industry
   architectures because major AI and cloud platforms are already
   exposing models to tools, APIs, enterprise workflows, code execution,
   operating-system actions, and physical robotics stacks.  The examples
   in this section are illustrative integration contexts only.  They do
   not assert that any named vendor has a security defect, lacks a
   particular control, or endorses this document.

   The common architectural trend is that model output is increasingly
   able to cause external state transitions.  As that authority grows,
   the security question shifts from whether a model may call a tool in
   principle to whether the exact consequential act crossing the final
   boundary is still permitted under current state.

12.1.  OpenAI: Long-Running Agents and Tool Execution

   OpenAI publicly describes its Agents API as infrastructure for
   building and running cloud agents that can manage context, use tools,
   coordinate subagents, work with files, run code, and persist
   intermediate state.  See Introducing the Agents API
   (https://openai.com/index/introducing-the-agents-api/).

   In such an architecture, execution-finality is potentially relevant
   below the agent harness: the model or harness may legitimately decide
   to invoke a tool, while a downstream finality sink can still verify
   the exact operation, current policy epoch, destination, replay state,
   and any execution-critical dependency before an irreversible external
   effect is released.

Das                       Expires 23 March 2027                [Page 23]
Internet-Draft        Agentic Effectuation Boundary       September 2026

12.2.  Anthropic: Autonomous Tool Use

   Anthropic describes current Claude models as capable of planning,
   using tools such as browsers and terminals, and running autonomously.
   See Introducing Claude Sonnet 5 (https://www.anthropic.com/news/
   claude-sonnet-5).

   Tool-use systems create a natural separation between model reasoning
   and external effect.  This document focuses on that separation: a
   tool call can be syntactically valid and originate from an authorized
   agent while the eventual side effect still requires exact-act
   binding, freshness, current-state validation, and anti-replay
   semantics.

12.3.  Microsoft: Autonomous Enterprise Agents

   Microsoft documents autonomous capabilities in Copilot Studio in
   which agents can react to events, make decisions, and execute tasks
   without waiting for an interactive user prompt.  See Design
   autonomous agent capabilities (https://learn.microsoft.com/en-us/
   microsoft-copilot-studio/guidance/autonomous-agents).

   Enterprise workflows frequently terminate in systems of record,
   ticketing systems, identity infrastructure, financial systems, or
   other stateful services.  The proposed finality boundary can be
   placed immediately before the mutation that carries the legally,
   financially, or operationally meaningful consequence, independently
   of which orchestration product generated the request.

12.4.  Google Cloud: Managed Agent Runtimes and Agent-to-Agent Systems

   Google Cloud documents Vertex AI Agent Engine as supporting
   deployment and operation of agents, including code execution, memory,
   and Agent-to-Agent protocol support.  See Vertex AI release notes
   (https://docs.cloud.google.com/vertex-ai/docs/release-notes).

   Multi-agent and multi-service systems make authority propagation
   especially important because the component that generated an intent
   may be several hops away from the component that performs the final
   effect.  The execution-finality model therefore treats propagated
   identity, authorization context, and attestation as inputs, while
   reserving the last act-specific decision for the effectuation
   boundary.

Das                       Expires 23 March 2027                [Page 24]
Internet-Draft        Agentic Effectuation Boundary       September 2026

12.5.  AWS: AgentCore and Enterprise Tool Connectivity

   AWS describes Amazon Bedrock AgentCore as a platform for building,
   deploying, and operating agents that take actions across tools and
   enterprise data with identity, access control, policy management,
   tool connectivity, session state, evaluation, and observability.  See
   Amazon Bedrock AgentCore overview (https://docs.aws.amazon.com/
   bedrock-agentcore/latest/devguide/what-is-bedrock-agentcore.html).

   Those controls are complementary to this document.  An execution-
   finality profile can sit at a payment adapter, deployment API, data-
   release gateway, or other consequence-producing resource and require
   a final exact-act and current-state check even when upstream runtime,
   identity, and policy controls all succeeded.

12.6.  NVIDIA: Robotics and Physical Actuation

   NVIDIA Isaac is an AI robotics development platform for autonomous
   mobile robots, robot arms, manipulators, and humanoids, with
   supporting motion-planning, perception, simulation, and deployment
   components.  See NVIDIA Isaac (https://developer.nvidia.com/isaac).

   Robotics makes the effectuation distinction concrete.  A planner can
   generate a valid trajectory while the physical environment changes
   before actuation.  The relevant sink may therefore be the command
   gate immediately before a motor controller, PLC, or robot control
   interface, where freshness, safety epoch, robot identity, command
   digest, and single-use release authority can be checked.

12.7.  Apple: On-Device Models, Tool Calling, and App Actions

   Apple documents its Foundation Models framework as supporting
   structured output and tool calling, while App Intents exposes app
   actions to system experiences including Siri and Apple Intelligence.
   See Foundation Models (https://developer.apple.com/documentation/
   foundationmodels) and Apple Intelligence
   (https://developer.apple.com/apple-intelligence/).

   On-device agentic systems are relevant because the final consequence
   may occur locally rather than at a cloud resource server.  A device-
   side finality sink can mediate file changes, communication, payment
   initiation, privacy-sensitive data release, peripheral control, or
   another local action without requiring the model itself to be the
   trusted enforcement component.

Das                       Expires 23 March 2027                [Page 25]
Internet-Draft        Agentic Effectuation Boundary       September 2026

12.8.  Cross-Vendor Interoperability Significance

   The industrial value of a standards-layer invariant becomes greater
   when the proposing agent, authorization server, workload runtime,
   tool provider, and effectuation system are operated by different
   vendors.  For example, a model from one provider may run in a cloud
   agent platform from another vendor, invoke a third-party enterprise
   tool, and ultimately request an action from a bank, industrial
   controller, robot, telecom network, or local device.

      Model Provider
           |
           v
      Agent Runtime
           |
           v
      Identity / OAuth / Workload Authorization
           |
           v
      Tool or Service Chain
           |
           v
      Third-Party Consequence Boundary
           |
           +--> money moves
           +--> data leaves
           +--> infrastructure changes
           +--> robot actuates

      Interoperability question:

      What machine-verifiable object and sink behavior allow the last
      component to verify the exact authorized consequence without
      trusting every upstream implementation choice?

12.9.  Why a Standards-Level Treatment Matters

   A single-vendor deployment can implement these properties as an
   internal engineering choice.  In a cross-vendor system, however,
   relying parties need an interoperable way to determine what was
   authorized, what exact act is being committed, which mutable state
   was required to remain current, whether the authority has already
   been consumed, and which component is responsible for the final
   check.

   This is the reason the topic is potentially relevant to the IETF: the
   proposed work is not intended to prescribe how OpenAI, Anthropic,
   Microsoft, Google, AWS, NVIDIA, Apple, or any other vendor should

Das                       Expires 23 March 2027                [Page 26]
Internet-Draft        Agentic Effectuation Boundary       September 2026

   build an agent.  It asks whether interoperable protocol artifacts and
   verification semantics are needed when independently operated systems
   exchange authority for actions whose consequences cannot safely be
   inferred from identity or bearer possession alone.

13.  Adversarial Test Matrix

    +======+================================+========================+
    | Test | Mutation or failure            | Expected result        |
    +======+================================+========================+
    | T1   | Change one execution-relevant  | Reject.                |
    |      | argument after authorization.  |                        |
    +------+--------------------------------+------------------------+
    | T2   | Replay an already consumed     | Reject without         |
    |      | handle.                        | duplicate effect.      |
    +------+--------------------------------+------------------------+
    | T3   | Advance policy or delegation   | Reject or require      |
    |      | epoch after handle issuance.   | reauthorization.       |
    +------+--------------------------------+------------------------+
    | T4   | Strip required lineage         | UNKNOWN; remain non-   |
    |      | metadata.                      | effective.             |
    +------+--------------------------------+------------------------+
    | T5   | Use a valid handle at a        | Reject.                |
    |      | different sink.                |                        |
    +------+--------------------------------+------------------------+
    | T6   | Retry after ambiguous network  | No duplicate effect.   |
    |      | timeout and committed effect.  |                        |
    +------+--------------------------------+------------------------+
    | T7   | Change a committed dependency  | Reject if that         |
    |      | after the final upstream read. | dependency is declared |
    |      |                                | execution-critical.    |
    +------+--------------------------------+------------------------+
    | T8   | Invoke an alternate            | Path considered non-   |
    |      | effectuation path that omits   | conformant or outside  |
    |      | finality verification.         | governed coverage.     |
    +------+--------------------------------+------------------------+
    | T9   | Change tool or mapping         | Reject or re-evaluate. |
    |      | revision with identical high-  |                        |
    |      | level request text.            |                        |
    +------+--------------------------------+------------------------+
    | T10  | Present valid attestation      | Reject the act while   |
    |      | evidence with an act outside   | preserving attestation |
    |      | authority.                     | result.                |
    +------+--------------------------------+------------------------+

                     Table 4: Minimum Negative Tests

Das                       Expires 23 March 2027                [Page 27]
Internet-Draft        Agentic Effectuation Boundary       September 2026

14.  Relationship to Current IETF Work

   OAuth deployments can provide delegated authorization, structured
   authorization detail, token sender constraint, and authorization
   context.  Execution-finality can consume those artifacts rather than
   duplicate them.  The distinguishing requirement is that the final
   effectuation component verifies the exact consequential act and any
   execution-critical current state.

   RATS and EAT can provide evidence about the platform, workload, or
   security state involved in issuing or consuming execution authority.
   Such evidence can become an input predicate without conflating
   attestation with transaction authorization.

   HTTP conditional requests provide a useful model for preventing a
   state-changing method from operating on a target representation that
   no longer matches an expected entity tag.  High-consequence agentic
   systems may additionally depend on non-HTTP state, multiple
   resources, policy generations, physical state, delegated authority,
   or cross-service prerequisites.

15.  Operational Considerations

   Not every agent action needs a high-assurance finality gate.  A
   deployment can define consequence classes and require stronger
   enforcement only for high-value, irreversible, safety-relevant, or
   externally binding operations.  This allows fast local verification
   on the hot path while slower logging, transparency, or anchoring
   mechanisms operate off the critical path.

   Implementers should explicitly document what the sink protects, which
   fields are part of the act digest, which states are revalidated, the
   lifetime of authority, crash recovery behavior, and known bypass
   paths.  An assertion of "atomic authorization" without a documented
   failure model is not independently testable.

16.  Privacy Considerations

   Execution handles should avoid carrying unnecessary personal or
   commercially sensitive information.  Where feasible, the handle can
   carry commitments or opaque references while the sink retrieves the
   minimum state required for verification.  Audit receipts should be
   designed to prove relevant facts without automatically becoming
   detailed activity-surveillance logs.

Das                       Expires 23 March 2027                [Page 28]
Internet-Draft        Agentic Effectuation Boundary       September 2026

17.  Security Considerations

   The finality sink becomes a high-value target.  Its key material,
   anti-replay state, policy epoch, canonicalization code, and commit
   adapter require protection commensurate with the consequence being
   governed.  A sink that can be bypassed, rolled back, or induced to
   canonicalize two different acts identically does not provide the
   claimed property.

   Denial of service is a deliberate tradeoff of fail-closed
   enforcement.  Deployments should distinguish states that truly
   require fresh verification from states that can be safely cached, and
   should define emergency procedures without creating an unaudited
   permanent bypass.

   Exact-act binding is only as strong as the chosen canonical form.
   Indirect effects, server-side defaults, wildcard expansion, schema
   upgrades, currency conversions, routing substitutions, or device-side
   transformations may need to be incorporated by digest, version, or
   explicit constraint.

18.  Open Questions for the IETF Community

   1.  Which existing token or capability formats are most suitable for
       carrying exact-act commitments without creating a new token
       family?

   2.  Which state belongs in a portable protocol object and which state
       should remain sink-local?

   3.  How should cross-service dependency commitments be represented
       without turning every action into a distributed transaction?

   4.  What minimal evidence allows a relying party or auditor to
       distinguish a deployment with atomic consume semantics from one
       that merely claims them?

   5.  How should effectuation-path coverage be described when a device
       has multiple legacy, emergency, local, or hardware control paths?

   6.  Can Transaction Tokens, RATS evidence, OAuth authorization
       details, or COSE objects be profiled to express this boundary
       without duplicating existing standards?

19.  IANA Considerations

   This document has no IANA actions.

Das                       Expires 23 March 2027                [Page 29]
Internet-Draft        Agentic Effectuation Boundary       September 2026

20.  Additional Public Technical Resources

   The following non-normative resources provide broader architectural
   background and runnable reference material.  Their inclusion does not
   make them part of the protocol requirements of this document.

   *  The Internet Solved Communication.  It Never Solved Authority.
      (https://zenodo.org/records/22082995)

   *  Execution-Finality Architecture for Machine-Generated Acts.
      (https://github.com/sangmdas/Execution-Finality-Architechture-for-
      AI-Machines-)

   *  tool_use Is Not invoke(): Runnable Agentic Tool-Call Reference
      Implementation.  (https://github.com/sangmdas/tool_use-Is-Not-
      invoke-Binding-Execution-Finality-to-Agentic-Tool-Call-Interfaces-
      and-MCP)

21.  Normative and Informative References

   [RFC8705]  Campbell, B., Bradley, J., Sakimura, N., and T.
              Lodderstedt, "OAuth 2.0 Mutual-TLS Client Authentication
              and Certificate-Bound Access Tokens", RFC 8705, February
              2020, <https://www.rfc-editor.org/info/rfc8705>.

   [RFC9110]  Fielding, R., Nottingham, M., and J. Reschke, "HTTP
              Semantics", RFC 9110, June 2022,
              <https://www.rfc-editor.org/info/rfc9110>.

   [RFC9334]  Birkholz, H., Thaler, D., Richardson, M., Smith, N., and
              W. Pan, "Remote ATtestation procedureS (RATS)
              Architecture", RFC 9334, January 2023,
              <https://www.rfc-editor.org/info/rfc9334>.

   [RFC9396]  Lodderstedt, T., Richer, J., and B. Campbell, "OAuth 2.0
              Rich Authorization Requests", RFC 9396, May 2023,
              <https://www.rfc-editor.org/info/rfc9396>.

   [RFC9449]  Fett, D., Campbell, B., Bradley, J., Lodderstedt, T.,
              Jones, M., and D. Waite, "OAuth 2.0 Demonstrating Proof of
              Possession (DPoP)", RFC 9449, September 2023,
              <https://www.rfc-editor.org/info/rfc9449>.

   [RFC9700]  Lodderstedt, T., Bradley, J., Labunets, A., and D. Fett,
              "Best Current Practice for OAuth 2.0 Security", RFC 9700,
              January 2025, <https://www.rfc-editor.org/info/rfc9700>.

Das                       Expires 23 March 2027                [Page 30]
Internet-Draft        Agentic Effectuation Boundary       September 2026

   [RFC9711]  Lundblade, L., Mandyam, G., O'Donoghue, J., and C.
              Wallace, "The Entity Attestation Token (EAT)", RFC 9711,
              April 2025, <https://www.rfc-editor.org/info/rfc9711>.

   [TXN-TOKENS]
              Tulshibagwale, A., Fletcher, G., and P. Kasselman,
              "Transaction Tokens", Work in Progress, Internet-Draft,
              draft-ietf-oauth-transaction-tokens-11, 30 July 2026,
              <https://datatracker.ietf.org/doc/draft-ietf-oauth-
              transaction-tokens/>.

Author's Address

   Sangam Das
   Independent Researcher
   Balasore
   Odisha
   India
   Email: info@sangamdas.com

Das                       Expires 23 March 2027                [Page 31]