Documented protection against cyberattacks, but no AI policy and no AI mandate
Overall rating “limited”. No deduction for controversies. The superscript numbers link directly to the source.
1 · Responsibility
Eaton names AI as a risk factor and calls for responsible use in its code of conduct, but has documented neither an AI policy nor an approval procedure for purchased AI.
What is documented
- The board's Innovation & Technology Committee reviews technology trends and risks of new technologies; it met twice in 20251 2
- The 2025 10-K treats AI as a separate risk factor, including data leakage to generative AI, accuracy, bias and regulation4
- The Code of Ethics calls for responsible use of AI and prohibits entering confidential information into unauthorised applications5
- An ethics help line is available around the clock, in several languages and anonymously; a data protection complaint channel exists5 7
The case against
AI governance not implemented
- Neither the board nor any committee has an explicitly defined responsibility for AI1
- No AI policy, no risk classification and no approval procedure for purchased AI such as Copilot
- No named person responsible for AI governance at group management level
No accountability on AI
- No AI metrics, no figures on AI training and no external audit of AI governance
- The Ethisphere ethics award concerns general compliance, not AI8
Sources for Responsibility (8)
- Eaton Corporation plc (SEC EDGAR): Proxy Statement (DEF 14A) 2026 · 13 March 2026 · primary source
- Eaton Corporation plc: Eaton 2025 Sustainability Report (PDF) · 2 July 2026 · secondary source
- Microsoft: Eaton – Microsoft 365 Copilot (customer story) · 19 November 2024 · secondary source
- Eaton Corporation plc (SEC EDGAR): Form 10-K for fiscal year 2025 · 26 February 2026 · primary source
- Eaton Corporation plc: Our Code: The Power of Integrity (Code of Ethics, English, publication BR090023EN) · 1 January 2026 · secondary source
- The Muse (job listing, Eaton): Specialist – AI Governance (Pune, Job ID Eaton-42882Pune) · 3 November 2026 · secondary source
- Eaton Corporation plc: Dispute resolution (data protection, UK country page) · 3 November 2026 · secondary source
- Business Wire (Eaton press release): Eaton Recognized for the 15th Time as One of the World's Most Ethical Companies · 20 March 2026 · secondary source
2 · Safety
Eaton documents a cyber programme based on NIST, a PSIRT with advisories and NREL test documentation for HiZ Protect, but field data and validation of the AI controls are missing.
What is documented
- Item 1C of the 10-K describes NIST CSF, monitoring, threat hunting, mandatory training, an incident response plan and supplier reviews5
- The PSIRT responds within two working days, grants safe harbour and states reporting deadlines under the EU Cyber Resilience Act8
- Advisories state CVE, CVSS and measures, including in one case without a patch due to end of support9 10
- NREL trained neural networks for HiZ Protect on Eaton test and grid data in simulated fault scenarios1
- Individual products are certified to UL 2900 and IEC 62443-4-26 7
The case against
AI security not documented
- For HiZ Protect, false alarm and failure rates, field evidence and information on human override are missing1
- No validation information is available for Brightlayer Energy and the Digital Energy Twin
- No AI-specific misuse monitoring and no usage restrictions documented
Sources for Safety (12)
- National Renewable Energy Laboratory (NREL): AI Could Snuff Out Wildfires One Power Line at a Time · 27 August 2025 · primary source
- Business Wire (Eaton press release): Eaton Announces Breakthrough AI Powered Innovation to Help Utilities Strengthen Wildfire Prevention Efforts · 24 March 2025 · secondary source
- 3BL Media (Eaton press release): Eaton Unveils Brightlayer Energy, an AI-Powered Energy Management and Optimization Software · 17 March 2026 · secondary source
- Business Wire (Eaton press release): Eaton Accelerates Transformation of Building and Data Center Infrastructure With Autodesk to Deliver AI Powered Digital Energy Twin and Software Tools · 15 September 2025 · secondary source
- Eaton Corporation plc (SEC EDGAR): Form 10-K for fiscal year 2025 · 26 February 2026 · primary source
- Eaton Corporation plc: Managing cybersecurity risks (Innovation Story) · 3 November 2026 · secondary source
- Security Magazine: Eaton achieves IEC and UL cybersecurity certifications for product development processes · 29 October 2020 · secondary source
- Eaton Corporation plc: Vulnerability Disclosure Policy · 3 November 2026 · primary source
- Eaton Corporation plc (PSIRT): Eaton Vulnerability Advisory ETN-VA-2025-1022 (xComfort ECI) · 22 December 2025 · primary source
- cvefeed.io (reproduction of NVD entry): CVE-2026-22618 – Eaton Intelligent Power Protector · 16 April 2026 · secondary source
- Eaton Corporation plc: Report an issue (Cybersecurity, UK country page) · 3 November 2026 · secondary source
- Eaton Corporation plc: Privacy, cookies and data protection (privacy notice, UK country page) · 1 March 2026 · secondary source
3 · Transparency
Eaton discloses AI use in customer service with human review, but for its AI products gives only accuracy claims without method, limits or system information.
What is documented
The case against
What is not disclosed
- Accuracy claims for HiZ Protect and Brightlayer Energy come from press releases without method or limits
- No labelling rules found for AI outputs in Brightlayer products
- No model or system cards and no transparency reports on AI features in products
Sources for Transparency (5)
- Eaton Corporation plc: Privacy, cookies and data protection (privacy notice, UK country page) · 1 March 2026 · secondary source
- Eaton Corporation plc (SEC EDGAR): Form 10-K for fiscal year 2025 · 26 February 2026 · primary source
- Business Wire (Eaton press release): Eaton Announces Breakthrough AI Powered Innovation to Help Utilities Strengthen Wildfire Prevention Efforts · 24 March 2025 · secondary source
- 3BL Media (Eaton press release): Eaton Unveils Brightlayer Energy, an AI-Powered Energy Management and Optimization Software · 17 March 2026 · secondary source
- National Renewable Energy Laboratory (NREL): AI Could Snuff Out Wildfires One Power Line at a Time · 27 August 2025 · primary source
4 · Why the harm rule does not apply
The harm rule does not apply: no AI-related judgment, fine or pattern of incidents was found during the period.
5 · Controversies (0)
No AI controversy found.
6 · What remains open
- Explicit board responsibility for AI and charter of the Innovation & Technology Committee
- Stand-alone AI policy, risk classification and approval procedure for purchased AI
- AI-specific tests, performance metrics and limits of HiZ Protect and Brightlayer
- Human override for automatically switching AI functions
- Wording of the legal proceedings disclosures in the 2025 10-K and the 10-Q for the third quarter of 2026
7 · How it was assessed
Cut-off date: 3 November 2026 · 40 searches, 30 fetches.
Primary documents read (9)
- Eaton Form 10-K FY2025 (up to character 100,000, incl. Item 1A and Item 1C)
- Eaton Proxy Statement (DEF 14A) 2026 (up to character 100,000)
- Eaton 2025 Sustainability Report (two sections, gap approx. pp. 40–59)
- Eaton Code of Ethics «Our Code: The Power of Integrity», January 2026 (English, Ukrainian)
- Eaton Vulnerability Disclosure Policy and Advisory ETN-VA-2025-1022
- Eaton Privacy Notice (as of March 2026) and dispute resolution page
- Form 8-K relating to the 2025 Annual General Meeting
- Form 10-Q Q3 2026 (up to Note 8)
- NREL report on high-impedance fault detection using neural networks
Research of 3 November 2026 under Globalance AI Footprint rulebook version 0