Hiroshima AI Reporting Framework

The Hiroshima AI Reporting Framework is open to organisations across the AI value chain, including developers, deployers, and providers of advanced AI systems. Participants are invited to report on their practices for managing risks and advancing trustworthy AI. Participation strengthens transparency, contributes to a shared understanding of organisational practices, and enables organisations to demonstrate their approaches to trustworthy AI.

Anthropic PBC: G7 Hiroshima AI Process (HAIP) Transparency Report

Publication date: Oct 2, 2026, 07:33 AM GMT+2

Reporting period: Q2 2026

Download

Section 0 - Your role in the AI system lifecycle

Responding as (please select all that apply):

Model developer (/provider)
Application developer (/provider)

If any of the fields above are selected, please indicate your organization size:

Large enterprise: 250 or more employees

Section 1 - Risk Identification and Evaluation

1. How does your organization define and/or classify different types of risks related to AI?¹

Maintain an internal taxonomy, classification system, and/or set of risk area definitions
Define enhanced safety measures for specific dangerous capability thresholds, that, once reached, require safeguards (e.g. capability threshold for dangerous Chemical, Biological, Radiological and Nuclear (CBRN), cybersecurity attacks, dangerous advanced autonomy capabilities, or other capabilities).
Align with specific external frameworks, standards, reports, and/or guidance documents such as taxonomies, classification systems, and/or sets of definitions for risk areas (e.g. the NIST AI Risk Management Framework, the ISO/IEC 42001, the OECD AI Principles, the EU General-Purpose AI Code of Practice, the International AI Safety Report). (please specify): ISO/IEC 42001:2023 (certified), NIST AI RMF (contributor)
Consult with external stakeholders (e.g., from academia, industry groups, civil society, or standard development organisations) when developing or updating risk taxonomies

How does your organization define thresholds at which severe risks, unless adequately mitigated, may be considered "unreasonable"? Please describe the framework(s) most relevant to your organization's role and context. If your organisation does not use formal thresholds, you may describe instead the policies, decision criteria, or governance processes you apply when assessing severe AI risks. You may also describe: • Specific examples of contexts in which your models or systems could, in your assessment, pose unreasonable or intolerable risk; • How mitigations are graduated by capability or propensity level; • How your organisation considers the specificities of different AI technologies, such as generative AI and agentic AI, including any specific risk assessment or governance mechanisms applied to agentic AI capabilities.

Anthropic's Responsible Scaling Policy (RSP), first published in 2023 and updated in July 2026, defines Capability Thresholds for certain catastrophic risks and proposed mitigations for each tier. Anthropic’s Frontier Compliance Framework, published in December 2025 and updated in July 2026, similarly defines risk tiers in four risk areas (CBRN, sabotage/loss of control, harmful manipulation, cyber offense) and evaluates the model against tiered risk levels per area. If a model reaches a higher tier, launch is gated: the risk owner documents the justification and mitigations for proceeding. Please see our Transparency Hub for more details on our Responsible Scaling Policy and Frontier Compliance Framework.

2. Both before and after deployment, what practices does your organization use to identify and evaluate vulnerabilities, incidents, emerging risks and misuse throughout the AI system lifecycle?

Internal red teaming³
Independent third-party red teaming
Adversarial testing to detect technical vulnerabilities
Publicly available benchmarks and measurement tools
Internal benchmarks, standard metrics (e.g., frequency, severity scores)
Structured techniques to explore the behavior or performance of a model and/or application, such as model elicitation or bottleneck assessments
Qualitative assessments (e.g., expert judgment, narrative scoring)
Specific risk identification practices for vulnerable groups
Impact assessments such as personal data protection assessment and fundamental rights impact assessments
Internal model, system, or application-independent research initiatives
Multilingual and cross-cultural evaluation, including collaboration with local ecosystems (e.g., national AI institutes, language-specific evaluation consortia) for under-represented languages and cultural contexts (please specify): As we mention on our Transparency Hub, we publish Multilingual benchmark results in our system cards
External model, system, or application-independent research initiatives
Vulnerability reports
Incident reports, including reports shared by other organizations
Automated monitoring / remote status reporting (e.g., telemetry pipelines, automated alerts when risk thresholds are breached)
Specific risk identification practices for agentic AI capabilities, e.g., autonomous action-taking, tool use, multi-step planning (please specify): We publish agentic safety evaluations in our system cards, which are released with each model. More details about our System Cards can be found on our Transparency Hub.

Are any national AI safety / security institutes of the AISI network involved?

Yes

Please briefly describe your organization’s decision-making process for deploying an AI model or system including: who has decision authority; what evidence and criteria are required (e.g., performance, safety/security, robustness, fairness, privacy, compliance, independent external testing, evaluation, validation and verification (TEVV) or assurance); contextual caveats or limitations required in interpreting risk evaluation metric results; what approvals/sign-offs are needed; and what safeguards (e.g., mitigations, monitoring, or phased release) must be in place to move into deployment.

Deployment decisions for frontier models are governed by our RSP and FCF. Before release we conduct evaluations to detect warning signs of increased catastrophic risk, threat modeling with external experts, internal and external red teaming (including with the UK AI Security Institute, US CAISI, and METR), and pre-deployment testing. Safeguards must be commensurate to assessed capabilities before a model moves to deployment.

We also publish Risk Reports every 3–6 months discussing the risks of our AI systems and our determinations about whether to continue development and deployment in light of those risks; these are shared with our Board of Directors and the Long-Term Benefit Trust and are subject to review by independent external parties. More information about these processes can be found on our Transparency Hub.

Following deployment, how do you use incident reports — including those shared by other organizations — to identify emerging issues, update risk assessments, and trigger corrective actions.

We update our Usage Policy and detection methods (including classifiers) based on how our models are being used in practice. Incident and vulnerability reports reach us through user and customer reporting, our responsible‑disclosure and bug‑bounty programs (including model‑safety bounties for jailbreaks), and automated post‑deployment monitoring. Our Threat Intelligence team identifies sophisticated misuse, including by correlating observed usage patterns with external threat data, and works within our Safeguards organization to improve defenses. Results feed back into model training, classifier updates, and enforcement strategy. We publish public threat-intelligence reports on identified misuse, and we participate in the Frontier Model Forum's information-sharing of frontier AI threats and vulnerabilities. More information about our incident reporting can be found in the Transparency Hub.

3. Does your organization have mechanisms to receive and/or share reports of risks, incidents or vulnerabilities by/with third parties?

Yes, both receiving and sharing

If yes, please indicate which stakeholders reporting mechanisms may be accessible to, recognising specifics will depend on the nature of the report and risk level. For each, indicate whether access is on a standing basis (Always), activated based on context (Ad hoc/as relevant), or not currently accessible.

	Always	Ad hoc / as relevant	Not currently accessible
Relevant employees (security, legal/compliance, etc.)	✓
Public authorities (e.g., public reporting offices for security incidents or data breaches)	✓
Supervisory authorities (e.g., data protection authorities)	✓
Industry partners or industry fora	✓
Users (end users, enterprise customers, developers using APIs)	✓
Business customers / deploying organizations (if different from “users”)	✓
Supply-chain partners (cloud, data providers, etc.)	✓
Academia / researchers	✓
Independent auditors	✓
Standards bodies / multi-stakeholder initiatives (OECD.AI, ISO/IEC, IEEE, etc.)	✓
Specific civil society organizations	✓
Media / general public (people or communities indirectly impacted)	✓
Other

Please describe how your organisation determines when and how reporting mechanisms are made accessible to different stakeholders (e.g. based on legal or regulatory requirements, internal policies, or risk level). Please also provide public URLs or details on how to submit reports. If none, write “Not applicable”

Information about our public reporting is available on our Transparency Hub.

How does your organization share relevant information as appropriate?

Publicly
Reporting to regulatory authorities as relevant
Information-sharing partnerships (with confidentiality agreements as appropriate)

Please describe what type of information is shared.

Information about our public reporting is available on our Transparency Hub.

4. If your organization has programs to incentivize risk(s) disclosure, incidents and systemic vulnerabilities, please share what type of programs you use?

Paid rewards (“bug bounty”)
Vulnerability Disclosure Programs (VDPs) with safe harbor / liability protection
Research grants, contracts, and hiring red teamers
Internal incentives and culture

Please provide public URLs and/or details on your incentive program(s). If none, write “Not applicable.”

anthropic.com/responsible-disclosure-policy · hackerone.com/anthropic-vdp · support.claude.com https://www.anthropic.com/transparency/voluntary-commitments

5. Does your organization use or contribute to the development of international technical standards, best practices or tools for the identification, assessment, and evaluation of risks?

Contributes and uses

If contributes or uses, please select all that apply:

ISO/IEC standards (e.g., ISO/IEC 27001, ISO/IEC 42001, ISO/IEC 23894, ISO/IEC 42005)
NIST AI Risk Management Framework (AI RMF)
Industry-led best practices or codes of conduct
General Purpose AI Code of Practice
Other international or regional standards (please specify): G7 Hiroshima Process Code of Conduct; Frontier AI Safety Commitments (Seoul); EU Code of Practice on Transparency of AI‑Generated Content (Art. 50); and C2PA content credentials

6. Does your organization collaborate with relevant stakeholders across sectors to assess and adopt risk mitigation measures to address risks that could have broad and potentially cascading socio-economic impacts? Please indicate which stakeholders you collaborate with, and whether collaboration is on a standing basis (Always-on), activated based on context (Ad hoc/as relevant), or there is no current collaboration.

	Always-on	Ad hoc / as relevant
Relevant employees (security, legal/compliance, etc.)	✓
Public authorities / regulators / AI safety institutes / other public sector bodies and inter-governmental coordination structures	✓
Industry partners or industry fora	✓
Users (end users, enterprise customers, developers using APIs)		✓
Business customers / deploying organizations (if different from “users”)		✓
Supply-chain partners (cloud, data providers, etc.)
Academia / researchers	✓
Independent auditors
Standards bodies / multi-stakeholder initiatives (OECD.AI; ISO/IEC; IEEE; AIID etc.)		✓
Independent civil society organizations (e.g., for evaluations, audits, or policy review)	✓
Media / general public (people or communities indirectly impacted)	✓
Other(s) (please add any other relevant stakeholder groups):
Not applicable

Please describe examples of how your organization collaborates. Examples may include: participating in multi-stakeholder or industry initiatives; partnering with academia, civil society, or the public sector; establishing internal or joint governance structures with external experts or in partnership with other efforts to share knowledge and jointly address risks.

More information about our Economic Impacts work can be found on our Transparency Hub.

Any further comments and implementation documentation.

No answer provided

Section 2 - Risk Management and Information Security

7. Across the AI system lifecycle, what steps does your organization take to manage risks and vulnerabilities? Please indicate which specific practices, tools or metrics your organization uses at each lifecycle stage. You may refer to tools or metrics present in the OECD.AI Catalogue of Tools and Metrics or suggest additions. Respondents may complete only the lifecycle stages relevant to their organisation's role. Stages that do not apply may be left blank.

Plan and design

Collect and process data

Build, integrate and interpret model

Verify and validate

Deploy

Operate and monitor

Please provide a consolidated description of your organisation's approach to risk management across the AI system lifecycle, including how the practices identified above operate together and any cross-cutting governance, review or escalation processes. Please also indicate how often these practices are reviewed and what types of events trigger reviews and updates. Where relevant, please distinguish between different types of AI systems (e.g. generative or agentic AI).

Risk management operates under two connected frameworks: the RSP (voluntary safety policy) and the FCF (compliance framework covering cyber offense, CBRN, AI sabotage/loss of control, and harmful manipulation, including model-weight protection and safety-incident response). More information about both can be found on our Transparency Hub.

Please describe any measures used to ensure data quality and fairness across the data collection and curation stage (e.g., dataset audits, filtering training data to remove harmful content or personal information, fine-tuning on curated data, use of moderation tools, assessment of limitations to model generalisability).

We undertake data preparation and cleaning processes to ensure that training data is of sufficient quality and appropriateness. We are adopting interventions to avoid ingesting CSAM, CSEM and NCII into our training datasets, including hash‑matching and reporting mechanisms with third parties to prevent CSAM ingestion and an NSFW filter applied to images from crawl sources. Our web crawler operates transparently, respects robots.txt signals, and does not access password‑protected or login‑gated pages, and we conduct due diligence on the training data we use.

8. What security measures are used in testing AI systems?

Isolation of staging environments (from production / operational environments)
Isolated and secure test environments
Network isolation from external access
Restrictions on outbound Internet or API access, including a list of permitted domains or APIs
Locally hosted testing environments
Controlled environments for red-teaming and adversarial testing
Containment measures for prompt-injection and jailbreak attempts
Controlled security vulnerability testing
Human-in-the-loop controls for higher-risk test cases
Rate limits and resource caps
Access controls, including roles-based permissions
Audit logging and monitoring
Disabling, restricting or simulating high-impact capabilities during testing
Use of synthetic, anonymized, or dummy data, as appropriate

8.A. Please indicate, if relevant, the team or role responsible for overseeing secure testing and whether this responsibility is shared across multiple teams and/or functions.

Our Security and Safeguards teams share responsibility for overseeing secure testing.

9. How does your organization protect intellectual property, including copyright-protected content?

9.A. Safeguards that promote lawful data use and transparency in AI training:

Maintain a copyright policy that addresses data sourcing, content provenance, safeguards, and organizational roles and responsibilities
Publishing “model cards” or contextual information on training data and model limitations
Maintain policies or practices to respect copyright holders’ preferences through, where appropriate/applicable, mechanisms like Robots.txt “do not train” tags or other machine-readable protocols developed through voluntary, industry-led standards processes
Apply provenance controls to guard against training data being sourced from websites or repositories that have been determined to engage in large-scale piracy
Refrain from the unlawful circumvention of technological protection measures in data sourcing
Conduct data cleaning, qualification, and/or tokenization activities as part of a transformative data preparation process
Engage in AI training that comprises a non-consumptive use of any underlying data
Consult and/or retain an intellectual property attorney to protect, enforce, and comply with, intellectual property rights.

9.B. Safeguards that reduce the risk of dissemination of outputs that may infringe copyright:

Restricting outputs that raise ethical, human rights, or legal concerns—such as non-consensual intimate imagery, unauthorized deepfakes, or outputs that are substantially similar to copyrighted works.
Filtering results using keyword or image-based restrictions to reduce risks of misuse

10. How do you protect privacy and prevent your systems from disclosing confidential or sensitive data across the AI lifecycle?

Make information about data privacy, practices, and controls accessible to a diverse set of stakeholders
Use security controls, including encryption at rest and in transit
Provide organizational customers with controls to configure and implement privacy choices (e.g., data retention, logging, access, or processing settings)

Describe key privacy measures and data-protection standards applied.

We comply with applicable data-protection laws and provide detailed disclosures about personal -data use and user controls via our Privacy Policy, Privacy Center, and Trust Center. More information about our approach to privacy can be found on our Privacy Center and Trust Center .

11. How does your organization assess cybersecurity risks and implement policies to enhance the cybersecurity of advanced AI systems across the AI system lifecycle? Respondents may complete the checkbox subsections in 11.A–11.D or describe their overall approach under 11.E.

11.A. Governance, policy and assessments

Regularly review security measures and update them in response to emerging risks, incidents, or major releases
Integrate cybersecurity AI risks into broader cybersecurity risk management activities
Train staff on cybersecurity best practices and AI-specific threats, including how to recognize and respond to misuse or unintended outcomes of advanced AI systems
Maintain vulnerability management processes (scanning, channels for external reporting, patching, remediation)
Implement incident response, disaster recovery, or business continuity plans, including escalation procedures for high-severity incidents
Maintain or update security policies aligned with recognized standards (e.g., NIST CSF, ISO/IEC 27001, SOC 2, CIS Critical Security Controls), and review them regularly in response to emerging risks or regulatory guidance
Perform threat modeling that includes AI-specific risks (e.g., data poisoning, prompt injection)
Conduct regular cybersecurity risk and vulnerability assessments tailored to AI models or systems

11.B. Technical and operational controls

Conduct security testing focused on AI-specific risks (e.g., red-teaming, adversarial testing, stress tests)
Monitor system activity and log access to detect anomalies or misuse
Conduct secure AI development and/or deployment practices
Restrict physical access to servers, GPUs, or devices hosting AI models, systems, and/or applications
Encrypt AI assets (datasets, models, configurations) in storage and in transit as appropriate
Apply multi-factor authentication (MFA) and role-based access controls for advanced AI systems

11.C. Protection of sensitive or high value AI assets

Require NDAs or confidentiality agreements for staff and contractors with access
Track and log access to sensitive AI resources
Use hardware security modules (HSMs) or trusted execution environments (TEEs) where feasible
Store model weights, datasets, and algorithms in secure, access-controlled environments
Limit access to proprietary AI assets based on role or need-to-know

11.D. Insider risk protection detections

Implement insider-risk escalation or response procedures
Provide staff training on cybersecurity best practices and AI-specific threats
Monitor for unauthorized internal access or suspicious activity
Conduct pre-employment screening or role-sensitive vetting

11.E. Please describe the main security measures your organization applies and indicate which are specific to particularly sensitive AI assets, where relevant.

Organization-wide measures. We operate a risk-based security program with periodic risk assessments reported to senior management, and hold ISO/IEC 27001:2022 and ISO/IEC 42001:2023 certifications alongside annual SOC 2 Type II examinations (reports via trust.anthropic.com). We maintain a threat-actor register and externally validated threat models spanning endpoint, supply-chain, physical, cloud, privilege-escalation and exfiltration vectors, and red-team AI-specific risks such as prompt injection. Baseline controls include multi-factor authentication and least-privilege, role-based access; managed-device access controls, endpoint detection and response, and application allowlisting on workstations; encryption at rest and in transit; network segmentation and data-loss controls; centralized logging with automated anomaly alerting; peer review and automated security analysis of code; continuous vulnerability scanning, risk-based remediation, annual external penetration testing, a public Responsible Disclosure Policy and bug bounties; layered physical security at our facilities and review of cloud service providers' independently audited data-center controls; and incident-response and business-continuity plans with escalation procedures for high-severity incidents, tested annually.

Measures specific to particularly sensitive AI assets. Our most sensitive AI assets are the weights of our most capable models; related training data and source code are restricted on a least-privilege, business-need basis. Weights additionally receive the ASL-3 protections our Responsible Scaling Policy (RSP) commits us to maintain or improve, designed to protect against most non-state attackers: multi-party, time-limited authorization using hardware security keys; device authorization limiting weight-system access to approved managed devices; egress bandwidth controls making large-scale exfiltration more time-consuming and detectable; cloud storage restrictions designed to prevent weights being written outside our security perimeter; network-source restrictions on privileged resources; additional review and commit-integrity requirements for critical repositories; and encryption, access logging and automated detection of unauthorized weight access or movement. Third-party partners hosting these models must implement equivalent security controls.

Personnel and insider risk (organization-wide). We apply pre-employment background checks where legally permissible; confidentiality agreements binding employees and contractor personnel before access is provisioned; security-awareness training at hire and annually, supplemented by AI-incident-response training for relevant staff; and a formal insider-risk program with internal monitoring, documented investigation and escalation procedures, and confidential and anonymous reporting channels, including RSP-noncompliance reporting with protection from retaliation.

Assurance. We publish RSP Risk Reports every three to six months and a Frontier Safety Roadmap of security goals; our Frontier Compliance Framework and "Activating ASL-3 Protections" report give more detail.

We run a dedicated insider-risk program under the security standards in our Responsible Scaling Policy: role-proportionate pre-employment verification, least-privilege access to model weights and sensitive assets with logging and monitoring, mandatory security training at hire and annually, and defined escalation channels to our Insider Risk and Security teams, including anonymous RSP noncompliance reporting. See our ASL-3 activation report and Trust Center.

12. What is your organization's vulnerability management process? Does your organization take actions to address identified risks and vulnerabilities, including in collaboration with other stakeholders?

Internal vulnerability disclosure receipt and tracking processes
Bug bounty or incentive programs for external reporters
Investigation of reported or identified risks and vulnerabilities
Development of patches or other mitigations for confirmed risks or vulnerabilities
Testing of patches or other mitigations
Scanning for and mitigation of similar risks or vulnerabilities
Collaborative vulnerability sharing (e.g., with partners, industry groups)
Reporting to regulators / authorities when relevant
Recognition of external finders of confirmed risks or vulnerabilities
Defined escalation criteria and follow-on actions for identified vulnerabilities (e.g., for external reporting or information sharing)

13. What post-deployment practices does your organization use to monitor and respond to vulnerabilities, incidents, emerging risks and misuse of advanced AI systems?

Conduct post-deployment monitoring
Perform root-cause analysis and apply lessons learned to improve future development and/or deployment practices
Monitoring and emergency controls specific to agentic AI systems (e.g., ability to interrupt autonomous action sequences, override tool-use permissions, or revoke delegated authority)

14. Do you develop or deploy agentic AI systems (systems that can autonomously plan/act or use tools)?⁴

Yes

14.A. If yes, please indicate which categories of controls your organization applies. Please select all.

Logging of agent actions
Monitoring (anomaly detection, multi-agent interaction monitoring)
Human oversight or interruptibility mechanisms (approval requirements for sensitive actions, escalation paths, interruption interfaces)
Constraints on the agent's action space (e.g., tool-use permissions, environment sandboxing, allowlists/blocklists of resources, scope limits per task)

Please provide any further comments and implementation documentation.

No answer provided

Section 3 - Transparency Reporting on Advanced AI Systems

15. Does your organization publish public reports and/or technical documentation related to the capabilities, limitations, and domains of appropriate and inappropriate use of advanced AI systems? Are these clear and understandable by non-technical audiences?

Yes

15.A. If yes, which of the following information is included in your organization’s publicly available documentation for any of your AI products?

Risk management policies
Summary of results from evaluations of safety, security, and societal risks including potential adverse impacts on individuals or groups (e.g. impacts on human rights, privacy or differential outcomes across user groups).
Plain-language or non-technical summaries
Technical documentation and instructions for use if relevant.
Capabilities, behaviours, and limitations, including domains of appropriate and inappropriate use or indicative thresholds associated with systemic risks.
Summary of results from red-teaming and other testing (e.g., adversarial evaluations, stress tests, and other pre-deployment safety assessments).

Please provide links or references showing where this information is made public.

anthropic.com/transparency · anthropic.com/system-cards · anthropic.com/news/responsible-scaling-policy-v3 · FCF on trust.anthropic.com · anthropic.com/legal/aup · docs.anthropic.com ·https://www-cdn.anthropic.com/f61d49fa5596956a5dec75fea0e973bf6a6a8378/Redacted%20Risk%20Report%20August%202026%20.pdf

15.B. If yes, how often are such reports usually updated?

At or before major product releases

16. How does your organization share information with a diverse set of stakeholders (other organizations, governments, civil society and academia, etc.) regarding the outcome of evaluations of risks and impacts related to an advanced AI system?

Public documentation
Protected information sharing
Regulatory reporting as relevant
Direct engagement with AI safety institutes, civil society or academic partners

Describe how evaluation results are shared and with which stakeholders (e.g. other organizations, governments, civil society and academia, etc). You may also indicate whether non-technical or plain-language summaries are provided for broader audiences.

Evaluation results are shared publicly through our system cards, Model Reports in our Transparency Hub, Risk Reports published under our Responsible Scaling Policy (with pilot external reviews by third parties), and research publications. Protected sharing occurs through pre-deployment testing with external organizations, the Frontier Model Forum's information-sharing arrangements, external red-team and evaluation partners, and vetted industry partners.

17. Does your organization publish public privacy policies addressing the use of personal data, user prompts, and/or the outputs of advanced AI systems?

Yes

If yes, provide links to privacy policies or notices covering personal data, user inputs, and AI-generated outputs.

18. How does your organization provide information about the sources of data used for the training of advanced AI systems, as appropriate, including information related to the sourcing of data annotation and enrichment?

Public documentation
Protected information sharing
Regulatory reporting as relevant

Please provide links to any publicly available information.

We provide information about our training data sources in our System Cards, Model Reports, Training Data Summaries, and Model Documentation Forms.

19. Does your organization demonstrate transparency related to advanced AI systems through any other methods?

Yes

If yes, describe additional method(s), including where relevant transparency measures related to AI agents

We publish Claude's constitution, describing the values and behaviors we aim for Claude to have; we publish our system prompts and release notes; we open-source safety evaluations and the dataset underpinning the Anthropic Economic Index; we publish interpretability research and usage-insights research; and we maintain public archives of prior Transparency Hub versions. Our research and findings related to the societal and economic impacts of advanced AI can be found in our Research Center.

20. Does your organization publish statistical insights derived from anonymised, aggregated usage data to advance understanding of AI adoption and its economic impacts in areas including occupations and tasks by region, sector and country? To this end, do you contribute to international cooperation efforts?

Yes

If yes, please provide links or describe the scope of insights published, including where relevant contributions to understanding AI adoption in specific sectors (e.g. public administration, education systems).

More information about our Economic Impacts work, including our Economic Index, is available at our Research Center.

Please estimate the total staff time and annual expenses associated with conducting compliance activities and include any further comments and implementation documentation (optional).

No answer provided

Section 4 - Organizational Governance, Incident Management, and Transparency

21. How has AI risk management been embedded in your organization's governance framework?

Executive oversight
Other (e.g., distributed / multi-institutional governance arrangement) (please specify): Read more about our Long Term Benefit Trust in our Transparency Hub

Please describe how AI risk management is integrated into governance structures, including: which roles or bodies have advisory versus decision-making authority; which roles or bodies have final accountability for AI risk-related decisions; and how often policies are reviewed or updated (e.g., following major incidents, regulatory changes, or material updates to systems).

Anthropic is a public benefit corporation. AI risk management is anchored in the RSP and FCF; Risk Reports on frontier-model risk and continued-deployment determinations are produced every 3–6 months, shared with the Board of Directors and the Long-Term Benefit Trust, and subject to independent external review in certain circumstances. The LTBT holds authority to elect Board members. The FCF and RSP are updated as our understanding of AI risk evolves (both most recently July 2026), and Transparency Hub reporting is updated regularly.

22. Are relevant staff trained on your organization's governance policies and risk management practices?

Yes

If yes, describe training methods (e.g., onboarding modules, regular sessions, and confidential channels to report concerns).

Employees have three AI-safety reporting channels — an emergency alerting and response system for potential harmful-use incidents, a general concern-sharing forum, and an anonymous channel for reporting potential RSP violations — with confidentiality and anti-retaliation protections.

23. Are steps taken to address reported incidents documented and maintained internally?

Yes

If yes, describe how incident records are kept and used for follow-up (e.g., incident logs, root-cause analysis reports, internal tracking systems). You may indicate whether and how you define escalation thresholds for when incidents must be surfaced to senior leadership or external parties.

Enforcement actions and appeals are tracked and reported in aggregate on the System Trust and Reporting page (banned accounts, appeals, overturns). More information about our reported incidents in our Transparency Hub

24. Does your organization share research and best practices on addressing or managing risk?

Yes

If yes, describe how research or good practices are shared (e.g., publications, conferences, collaborations with other sectors).

We regularly publish safety, interpretability, alignment, societal-impact, and frontier red-teaming research; we publish public threat-intelligence reports, open-source safety evaluations, and share best practices through the Frontier Model Forum.

25. Does your organization use international technical standards or best practices for AI risk management and governance policies?

Yes

If yes, please indicate which standards or frameworks apply:

Industry-led best practices or codes of conduct
OECD AI Principles
NIST AI Risk Management Framework (AI RMF)
ISO/IEC standards (e.g., ISO/IEC 27001, ISO/IEC 42001)

If yes, describe which international technical standards or best practices your organization uses.

We achieved ISO/IEC 42001:2023 certification for our AI management system — the first international standard for AI governance. We use and contribute to the NIST AI RMF. Our voluntary-commitment frameworks (G7 Hiroshima Code of Conduct, Seoul Frontier AI Safety Commitments, and others listed on our Transparency Hub) structure our public reporting.

Any further comments and implementation documentation.

No answer provided

Section 5 - Content Authentication & Provenance Mechanisms

26. What mechanisms, if any, does your organization put in place to allow end-users, where possible and appropriate, to know when they are interacting with an advanced AI system developed by your organization?

Documentation or FAQs explaining when and how AI is used in the system
Access to provenance information (e.g., model version, content authenticity signals)
Disclosure to end-users when an AI system is acting autonomously or executing actions on their behalf (e.g., making purchases, sending communications, modifying files)

Provide any additional details or context on how or when users are informed they are interacting with advanced AI systems (e.g. if such mechanisms are planned, in testing, or applied only to certain products).

No answer provided

27. Does your organization use content provenance detection, labeling or watermarking mechanisms that enable users to identify content generated by advanced AI systems?

Yes

27.A. If yes, how?

No answer provided

27.B. If yes, does your organization use international technical standards or best practices when developing or implementing content provenance?

Yes

27.C. Does your organization collaborate on or invest in research to advance content authentication and provenance?

Yes

Describe how provenance mechanisms are implemented and which standards or tools are used, and any collaborations or investments to improve content authentication (e.g., C2PA credentials, joint R&D projects, open-source initiatives, standard-setting groups). You may also indicate whether these mechanisms are tested for robustness against removal, tampering, or circumvention.

No answer provided

Any further comments and implementation documentation.

Anthropic embeds an imperceptible watermark in text generated by supported Claude models launched on or after August 2, 2026, applied globally at the model level across all Claude products and cloud partners. Generated files (.svg, .png, .jpg) receive signed C2PA provenance metadata where supported, which helps users to detect when content has been processed by Claude. More detail is available in our blog post.

Section 6 - Research & Investment to Advance AI Safety & Mitigate Societal Risks

28. Does your organization advance research and investment in the following areas?

Robustness or reliability
Transparency
Explainability or interpretability
System performance and outcomes (e.g. accuracy, consistency)
Information integrity
Data quality and representativeness
Safety
Security

If yes, please describe key research or investment initiatives in the areas selected above (e.g., dedicated safety teams, funding of research programs, contributions to open-source safety tools, publishing research, partnerships and collaborative research initiatives). You may also indicate whether related outputs are publicly shared, shared with partners or within consortia or not publicly disclosed.

We have dedicated research teams including Alignment Science (alignment.anthropic.com), the Frontier Red Team (red.anthropic.com), and interpretability (transformer-circuits.pub). We fund third-party evaluations through a dedicated initiative, commission external red-teaming and Policy Vulnerability Testing, operate model-safety bug bounties, and publish safety research including on Constitutional AI and Constitutional Classifiers. Research findings are publicly shared via our research blogs and publications.

29. Does your organization pursue research or investment related to socio-economic or environmental considerations in the development and deployment of AI?

Socio-economic considerations (e.g., AI adoption and productivity, workforce transitions, skills development, access to AI tools)
Environmental considerations (e.g., energy use, computational efficiency, resource use)

If yes, please provide examples of relevant research or investment initiatives for the areas selected above.

See https://www.anthropic.com/research

Any further comments and implementation documentation.

No answer provided

Section 7 - Advancing Human and Global Interests

30. Does your organization support any digital literacy, education or training initiatives to improve user awareness and/or help people understand the nature, capabilities and limitations of advanced AI systems?

Yes

If yes, describe any digital literacy or training initiatives your organization supports (e.g., user education, public AI-education programs, school or university partnerships, community workshops). You may also indicate whether these efforts include or prioritise underserved or underrepresented communities.

We provide information about our approach to AI literacy in our Transparency Hub.

31. Does your organization pursue AI projects that aim to deliver societal or public benefits, including through partnerships with civil society and community groups?

Yes

If yes, please provide examples, including any collaborations with civil society, community groups, or efforts to address broader societal challenges.

We provide more details on our Public Benefit work in our Transparency Hub.

Any further comments and implementation documentation.

No answer provided

Back to reports