AI Preferences L. Peiyuan Internet-Draft Condé Nast Intended status: Experimental 8 April 2025 Expires: 10 October 2025 Protocol for Basic Automation Control draft-liao-aipref-autoctl-core-00 Abstract This document specifies the core automation-preferences.txt protocol, a machine-readable document that defines server-side automation permissions with a focus on essential controls for AI and automation use cases. Unlike the traditional robots.txt, which governs crawling, this protocol addresses a broader range of automation activities while maintaining simplicity for initial implementation. It defines the fundamental file format, policy declarations, HTTP method restrictions, and purpose requirements. Advanced features are addressed in a separate extension specification. About This Document This note is to be removed before publishing as an RFC. The latest revision of this draft can be found at https://datatracker.ietf.org/doc/draft-liao-aipref-autoctl-core/. Status information for this document may be found at https://datatracker.ietf.org/doc/draft-liao-aipref-autoctl-core/. Discussion of this document takes place on the AI Preferences Working Group mailing list (mailto:ai-control@ietf.org), which is archived at https://mailarchive.ietf.org/arch/browse/ai-control/. Subscribe at https://www.ietf.org/mailman/listinfo/ai-control/. 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/. Peiyuan Expires 10 October 2025 [Page 1] Internet-Draft aipref-autoctl April 2025 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." This Internet-Draft will expire on 10 October 2025. Copyright Notice Copyright (c) 2025 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 . . . . . . . . . . . . . . . . . . . . . . . . 3 1.1. Applicability . . . . . . . . . . . . . . . . . . . . . . 3 1.2. Relationship to Extension Specification . . . . . . . . . 3 2. Conventions and Definitions . . . . . . . . . . . . . . . . . 3 3. Protocol Specification . . . . . . . . . . . . . . . . . . . 4 3.1. File Location and Format . . . . . . . . . . . . . . . . 4 3.2. Automation Policy Declaration . . . . . . . . . . . . . . 4 3.3. HTTP Method Restrictions . . . . . . . . . . . . . . . . 5 3.4. Purpose Declaration . . . . . . . . . . . . . . . . . . . 5 3.5. Scope and Applicability . . . . . . . . . . . . . . . . . 6 4. Extension Mechanism . . . . . . . . . . . . . . . . . . . . . 7 5. Implementation and Enforcement . . . . . . . . . . . . . . . 7 6. Security Considerations . . . . . . . . . . . . . . . . . . . 8 7. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 8 8. References . . . . . . . . . . . . . . . . . . . . . . . . . 8 8.1. Normative References . . . . . . . . . . . . . . . . . . 8 8.2. Informative References . . . . . . . . . . . . . . . . . 9 Sample automation-preferences.txt File . . . . . . . . . . . . . 9 Author's Address . . . . . . . . . . . . . . . . . . . . . . . . 9 Peiyuan Expires 10 October 2025 [Page 2] Internet-Draft aipref-autoctl April 2025 1. Introduction The evolution of web automation has outpaced the capabilities of existing standards such as robots.txt, which only provides for crawler permissions. This document introduces the core automation- preferences.txt protocol that enables server operators to explicitly define basic policies governing automated interactions. These core policies cover fundamental aspects of automation control, including HTTP method restrictions and automation purpose declarations. The automation-preferences.txt protocol is designed with extensibility in mind, allowing future enhancements while maintaining backward compatibility. This specification focuses on essential controls to facilitate initial adoption and implementation, with more advanced features defined in a separate extension document. 1.1. Applicability The automation-preferences.txt protocol applies to automated systems interacting with web servers, especially those driven by foundation models or other types of advanced AI models. It is designed to benefit both content owners-by allowing them to specify acceptable automation behaviors-and developers of automated systems, who can use these directives to ensure compliance. 1.2. Relationship to Extension Specification This document defines the core functionality of the automation- preferences.txt protocol. A separate document, "Protocol Extension for Advanced AI Automation Control," extends this core specification with additional directives and capabilities. The extension specification builds upon this core document without modifying its requirements, providing a path for progressive implementation. Implementations conforming to only this core specification are considered compliant with the automation-preferences.txt protocol. The extension specification defines optional enhancements that may be implemented once the core functionality is established. 2. Conventions and Definitions The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in BCP 14 [RFC2119] [RFC8174] when, and only when, they appear in all capitals, as shown here. The following terms are used in this document: Peiyuan Expires 10 October 2025 [Page 3] Internet-Draft aipref-autoctl April 2025 * Automation: Programmatic interactions with a web server. * State-changing requests: HTTP methods that alter server state (e.g., POST, PUT, DELETE, PATCH). * Automation policy: A set of rules defining permitted and prohibited automated behaviors for a specific scope. * Automation purpose: The declared intent or use case for which automation is being performed. All terminology defined in this core specification applies to the extension specification without redefinition. The extension specification may introduce additional terms for concepts not covered in this document. 3. Protocol Specification 3.1. File Location and Format The automation-preferences.txt file MUST be hosted at the root of the domain, in the same manner as robots.txt. The file is structured as a series of key-value pairs that specify automation permissions. The file MUST be served with the text/plain MIME type. Lines beginning with the hash symbol (#) are considered comments and MUST be ignored by parsers. Each directive consists of a field name, followed by a colon, followed by a value. Multiple values MAY be separated by commas. Parsers MUST silently ignore any directives they do not recognize. This enables future extensions to add new capabilities without breaking compatibility with existing implementations. 3.2. Automation Policy Declaration A top-level directive, AutomationPolicy, indicates the overall stance of the server regarding automated interactions and state-changing requests. The following values are defined: * *open*: Automation is generally permitted with few restrictions. * *limited*: Automation is permitted with specific restrictions. * *strict*: Automation is heavily restricted or prohibited except in specific circumstances. Peiyuan Expires 10 October 2025 [Page 4] Internet-Draft aipref-autoctl April 2025 If the AutomationPolicy directive is not present, clients SHOULD assume a default value of "limited". Example: AutomationPolicy: limited Figure 1 3.3. HTTP Method Restrictions The protocol MUST explicitly list allowed and disallowed HTTP methods using the AllowedMethods and DisallowedMethods directives. Typically, GET and HEAD are permitted while methods such as POST, PUT, DELETE, and PATCH are disallowed for automated processing. If no HTTP method directives are specified, clients SHOULD assume that only GET and HEAD methods are permitted. Example: AllowedMethods: GET, HEAD DisallowedMethods: POST, PUT, DELETE, PATCH Figure 2 3.4. Purpose Declaration The protocol enables automation clients to declare their intended usage purposes. This is managed using the following directives: * RequireAutomationPurpose: Boolean value indicating whether clients must declare a purpose for automation. * AllowedPurposes: Comma-separated list of permitted purposes using standardized vocabulary terms. * DisallowedPurposes: Comma-separated list of prohibited purposes. The specific vocabulary terms for automation purposes are intentionally not defined in this protocol specification. Instead, this protocol provides a mechanism for expressing allowed and disallowed purposes, which can be populated with terms from any widely accepted vocabulary standard. This approach ensures that the protocol remains flexible and can adapt to evolving vocabulary standards while maintaining the essential structure for purpose declarations. Peiyuan Expires 10 October 2025 [Page 5] Internet-Draft aipref-autoctl April 2025 Example: RequireAutomationPurpose: true AllowedPurposes: [PLACEHOLDER_PURPOSE1], [PLACEHOLDER_PURPOSE2] DisallowedPurposes: [PLACEHOLDER_PURPOSE3] Figure 3 3.5. Scope and Applicability In a manner similar to robots.txt, the automation-preferences.txt file is divided into groups, each of which applies to a specific subset of content. Each group begins with one or more scope directives that define the target of the preferences. The following directives MAY be used within a group: * Scope: Specifies the URL pattern (e.g., /admin/) to which the group applies. Wildcards MAY be used to indicate variable components of the URL. * Host: Specifies a subdomain or host. If present, the group applies only to the indicated subdomain; if omitted, the group is assumed to apply to the entire host. Groups are processed in order of specificity. When multiple groups match a given request, the group with the longest matching Scope directive SHALL take precedence. Example: # Group 1: Applies to the entire site Host: example.com Scope: / AutomationPolicy: limited AllowedMethods: GET, HEAD DisallowedMethods: POST, PUT, DELETE, PATCH # Group 2: Specific preferences for the /admin/ path Host: example.com Scope: /admin/ AutomationPolicy: strict AllowedMethods: GET DisallowedMethods: POST, PUT, DELETE, PATCH Figure 4 Peiyuan Expires 10 October 2025 [Page 6] Internet-Draft aipref-autoctl April 2025 4. Extension Mechanism The automation-preferences.txt protocol is designed for extensibility. The protocol defines a forward-compatible approach where implementations: * MUST process all recognized directives according to this specification * MUST silently ignore any unrecognized directives * MUST NOT fail or produce errors when encountering extended directives This enables future extensions to add new capabilities while maintaining compatibility with implementations of this core specification. Extended directives may include: * Rate limiting controls * Automation technology restrictions * API and XHR permissions * Session requirements * Asset-level annotation methods 5. Implementation and Enforcement Servers implementing this protocol SHOULD: * Verify incoming requests against the rules specified in automation-preferences.txt. * Respond with appropriate HTTP status codes (e.g., 403 Forbidden) for non-compliant requests. * Include HTTP headers that reference the applicable automation policy for transparency. Clients consuming this protocol SHOULD: * Fetch and parse the automation-preferences.txt file before performing automated operations. * Honor the HTTP method restrictions specified in the file. Peiyuan Expires 10 October 2025 [Page 7] Internet-Draft aipref-autoctl April 2025 * Declare their automation purpose when required. * Respect the scope directives when performing operations on different paths. Implementations SHOULD cache the automation-preferences.txt file to reduce server load, but SHOULD NOT cache it for longer than 24 hours to ensure timely policy updates. 6. Security Considerations The use of machine-readable automation policies introduces security considerations that must be addressed by implementations: * Parsing of automation-preferences.txt MUST be performed securely to prevent vulnerabilities such as buffer overruns and denial-of- service attacks. * Care SHOULD be taken to avoid exposing sensitive policy details that could be exploited by adversaries. * The protocol does not provide authentication or cryptographic verification mechanisms for the file content. Servers SHOULD ensure the file is served via secure connections to prevent tampering. * The protocol does not enforce client compliance; it relies on good-faith adherence by automation providers. Servers SHOULD implement additional detection and enforcement mechanisms as needed. 7. IANA Considerations This document has no IANA actions. 8. References 8.1. Normative References [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, DOI 10.17487/RFC2119, March 1997, . [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, May 2017, . Peiyuan Expires 10 October 2025 [Page 8] Internet-Draft aipref-autoctl April 2025 8.2. Informative References [RFC9309] Koster, M., Illyes, G., Zeller, H., and L. Sassman, "Robots Exclusion Protocol", RFC 9309, DOI 10.17487/RFC9309, September 2022, . Sample automation-preferences.txt File The following is an example of a automation-preferences.txt file that adheres to this specification: # Automation preferences for example.com # Version: 1.0 # Last updated: 2025-04-08 # Group 1: Applies to the entire site Host: example.com Scope: / AutomationPolicy: limited AllowedMethods: GET, HEAD DisallowedMethods: POST, PUT, DELETE, PATCH RequireAutomationPurpose: true AllowedPurposes: [PLACEHOLDER_PURPOSE1], [PLACEHOLDER_PURPOSE2] DisallowedPurposes: [PLACEHOLDER_PURPOSE3] ContactEmail: automation-policy@example.com # Group 2: Specific preferences for the /admin/ path Host: example.com Scope: /admin/ AutomationPolicy: strict AllowedMethods: GET DisallowedMethods: POST, PUT, DELETE, PATCH AllowedPurposes: [PLACEHOLDER_PURPOSE1] DisallowedPurposes: [PLACEHOLDER_PURPOSE2], [PLACEHOLDER_PURPOSE3] Figure 5 Author's Address Liao Peiyuan Condé Nast United States of America Email: peiyuan_liao@condenast.com Peiyuan Expires 10 October 2025 [Page 9]