AI oversight on the board and documented data security, but no published AI rules
Overall rating “limited”. No deduction for controversies. The superscript numbers link directly to the source.
1 · Responsibility
Travelers documents board oversight, training on AI risks and a procedure for its recruitment tool, but a published AI policy and an approval procedure are missing, hence “limited”.
What is documented
- In 2025 the board reviewed the AI strategy and AI governance strategies and received training on AI and agentic AI1
- According to the 2026 proxy, the use of AI in underwriting and pricing was a topic of shareholder engagement; one director brings AI experience2
- The 10-K names targeted training on AI-related risks9
- For the recruitment tool, how it works, review by recruiters and an opt-out without disadvantage are described5
- An independently operated ethics helpline reports quarterly to the Audit Committee10 4
The case against
No published AI rules
- An AI policy with responsibilities, review duties or prohibitions and an approval procedure for AI are not published
- No risk classification, named AI committee or external AI audits were found
- No committee is explicitly responsible for AI; the Risk Committee covers IT and cybersecurity2
Sources for Responsibility (12)
- The Travelers Companies, Inc.: Travelers Sustainability Report 2025 (PDF) · 6 July 2026 · primary source
- The Travelers Companies, Inc. / U.S. SEC: Definitive Proxy Statement 2026 (DEF 14A) · 6 April 2026 · primary source
- Fortune (via inkl): Why insurance giant Travelers' CTO is placing fewer, bigger bets on AI · 2026 · secondary source
- The Travelers Companies, Inc.: Sustainability – Ethics & Responsible Business Practices (web page) · 3 November 2026 · secondary source
- The Travelers Companies, Inc. (Careers): AI in Recruiting · 22 October 2026 · primary source
- Connecticut Insurance Department: Bulletin MC-25 – The Use of Artificial Intelligence Systems by Insurers · 26 February 2024 · secondary source
- Faegre Drinker: Colorado Division of Insurance Expands AI-Related Governance and Risk Management Obligations for Insurers · 2 September 2025 · secondary source
- New York State Department of Financial Services: Insurance Circular Letter No. 7 (2024) · 11 July 2024 · secondary source
- U.S. SEC (XBRL section of 10-K 2025): R33 – Cybersecurity Risk Management and Strategy Disclosure (Item 1C) · 12 February 2026 · primary source
- The Travelers Companies, Inc.: Code of Business Conduct and Ethics (version 2026-05-22, CP-16659 Rev. 5-26) · 22 May 2026 · secondary source
- The Travelers Companies, Inc.: California Consumer Privacy Act Disclosure (as of 2026-02-13) · 13 February 2026 · primary source
- Evident Insights: Evident AI Index for Insurance, 2026 Edition · 16 June 2026 · secondary source
2 · Safety
Travelers describes a security programme under ISO 27001 and NIST CSF with incident exercises, but authorities found shortcomings in 2024, and there are only own statements on testing the AI models.
What is documented
- The 10-K describes MFA, encryption, data loss prevention, penetration testing and reviews by auditors and the supervisor7
- An incident response framework with tabletop exercises and participation in FS-ISAC are documented7
- Cyber incidents are reported quarterly to the Risk Committee; incident playbooks exist4
- The New York settlement imposes requirements on authentication, logging and response8 9
- According to the company, callers using the AI voice assistant can switch to specialists at any time1
The case against
Officially established shortcomings
Sources for Safety (9)
- Business Wire / The Travelers Companies, Inc.: Travelers Launches Industry Leading Agentic AI Claim Assistant Developed with OpenAI · 18 February 2026 · secondary source
- The Motley Fool (transcript): Travelers (TRV) Q4 2025 Earnings Call Transcript · 21 January 2026 · secondary source
- Business Wire / The Travelers Companies, Inc.: Travelers Publishes Its 2025 Sustainability Report · 6 July 2026 · secondary source
- The Travelers Companies, Inc.: Travelers Sustainability Report 2025 (PDF) · 6 July 2026 · primary source
- KUT 90.5 (Audrey McGlinchy): In Texas, insurers are watching your home from above. It could cost you coverage. · 13 May 2025 · secondary source
- AI Incident Database: Entity Travelers Insurance – Incident 1083 · 13 May 2025 · secondary source
- U.S. SEC (XBRL section of 10-K 2025): R33 – Cybersecurity Risk Management and Strategy Disclosure (Item 1C) · 12 February 2026 · primary source
- New York State Office of the Attorney General: Attorney General James and DFS Superintendent Harris Secure $11.3 Million from Auto Insurance Companies over Data Breaches · 25 November 2024 · primary source
- New York State Department of Financial Services: Press Release 2024-11-25 – GEICO and Travelers · 25 November 2024 · primary source
3 · Transparency
For claims, underwriting and customer dialogue there are only general descriptions; whether the AI voice assistant is recognisable as an AI is not documented.
What is documented
The case against
Little information for those affected
No system documentation
- System documentation for the voice assistant and the internal AI platform is missing
- External data sources of the underwriting models are not disclosed
Sources for Transparency (8)
- Business Wire / The Travelers Companies, Inc.: Travelers Launches Industry Leading Agentic AI Claim Assistant Developed with OpenAI · 18 February 2026 · secondary source
- The Travelers Companies, Inc.: Privacy Statements (Online Privacy Statement, as of 2025-04-09) · 9 April 2025 · secondary source
- The Travelers Companies, Inc. (Careers): AI in Recruiting · 22 October 2026 · primary source
- Insurance Journal: AI Claim Assistant Now Taking Auto Damage Claims Calls at Travelers · 19 February 2026 · secondary source
- KUT 90.5 (Audrey McGlinchy): In Texas, insurers are watching your home from above. It could cost you coverage. · 13 May 2025 · secondary source
- The Travelers Companies, Inc.: Travelers Sustainability Report 2025 (PDF) · 6 July 2026 · primary source
- Hartford Business Journal: Travelers deploying AI agents for auto claims · February 2026 · secondary source
- The Travelers Companies, Inc.: California Consumer Privacy Act Disclosure (as of 2026-02-13) · 13 February 2026 · primary source
4 · Why the harm rule does not apply
The harm rule does not apply: the 2024 New York fine concerns a data breach with no AI link, and no pattern of serious AI incidents is documented.
5 · Controversies (3)
- minorA homeowner in Texas received a non-renewal warning based on aerial images; whether AI generated the flag is not documented. · 13 May 2025 · unklar
- minorA class action over the same data breach ended in a court-approved settlement; no AI link. · 5 February 2025 · Vergleich
- minorThe New York Attorney General and DFS fined Travelers in 2024 because driver's licence numbers were extracted in 2021 via an agent portal without MFA; no AI link. · 25 November 2024 · abgeschlossen
6 · What remains open
- Published AI policy, risk classification and approval procedure for AI
- Validation, monitoring and error rates of the AI models in claims and underwriting
- Labelling of the AI voice assistant towards callers
- Compliance with the requirements of Connecticut, Colorado and New York on AI in insurance
- Risk factors and legal proceedings in the 2025 10-K and the second half of the 2025 sustainability report could not be read
7 · How it was assessed
Cut-off date: 3 November 2026 · 41 searches, 38 fetches.
Primary documents read (10)
- Form 10-K Financial Year 2025 (Item 1, in HTML and PDF, two parts each)
- 10-K 2025 XBRL R33 (Item 1C Cybersecurity) and R25 (Contingencies)
- Definitive Proxy Statement 2026 (two parts, up to character 119,254)
- Definitive Proxy Statement 2024 (up to character 100,000)
- Form 8-K Voting Results 2026-05-20
- Sustainability Report 2025 (pp. 1–53, two parts)
- Code of Business Conduct and Ethics, version 2026-05-22
- AI in Recruiting, Privacy Statements and CCPA Disclosure
- Press releases from NY Attorney General and NY DFS dated 2024-11-25
- Connecticut Bulletin MC-25, NY Circular Letter No. 7, NAIC Notice on the Model Bulletin
Research of 3 November 2026 under Globalance AI Footprint rulebook version 0