AI bodies described only by the company itself, product security partly independently certified
Overall rating “limited”. No deduction for controversies. The superscript numbers link directly to the source.
1 · Responsibility
The Responsible AI Committee, RAI Office and risk framework are described only in blogs and interviews, and no AI mandate of the board is documented; what is documented is the agreement with the trade unions in France, published by third parties.
What is documented
- On 2026-05-22 Schneider and the trade unions in France unanimously signed an agreement establishing advance review of AI projects and human oversight9 10
- The committee of 15 trained employee representatives also monitors psychosocial risks; the agreement runs until the end of 20289 10
- According to the company, a Responsible AI Committee of those responsible for AI, products, security, data, legal and compliance reviews critical applications2 3
- The described risk framework provides for risk-based review levels from the design phase and alignment with the EU AI Act2 3
- A Chief AI Officer reports to the Chief Digital Officer4 6
The case against
Mostly self-description
Sources for Responsibility (13)
- Schneider Electric SE: Integrated Report 2025 · 3 November 2026 · secondary source
- Schneider Electric (Blog, Nicolas Rambaud): Cyber governance and AI: Balancing opportunity and risk through trustworthy AI · 2 September 2025 · secondary source
- Schneider Electric (Blog/Podcast «AI at Scale»): Where does AI governance stop? – Ania Kaci: AI governance decoded · 21 April 2025 · secondary source
- CDO Magazine: Schneider Electric CAIO: We avoid treating AI as innovation (Part 2) · 17 March 2026 · secondary source
- Schneider Electric SE: Board study committees · 29 July 2026 · secondary source
- Europäische Kommission (Shaping Europe's digital future): Philippe Rambach (speaker profile, Apply AI Summit) · 14 September 2026 · secondary source
- The Muse (job advertisement, Schneider Electric): Principal Responsible AI Operationalisation Officer · 3 November 2026 · secondary source
- Schneider Electric SE: Schneider Electric drives Generative AI productivity and sustainability solutions by integrating Microsoft Azure OpenAI · 13 November 2023 · secondary source
- industriAll European Trade Union: Schneider pioneers social dialogue for AI and the green transition · 30 June 2026 · primary source
- Mind RH (mind.eu.com): France: how Schneider Electric is developing social dialogue tailored to AI and the green transition · 22 May 2026 · primary source
- PYMNTS: Schneider Electric Deploys AI Across 100 Use Cases in Industrial Operations · 1 April 2026 · secondary source
- Schneider Electric SE: 2023 Trust Report · 3 November 2026 · secondary source
- Energy Digital (Energy Magazine, Juni 2026): Energy Magazine June 2026, p. 150 (interview with Philippe Rambach) · 1 June 2026 · secondary source
2 · Safety
A TÜV certificate under IEC 62443-4-1, vulnerability advisories published by CISA and a disclosure policy document product security; AI-specific tests are only self-described, and two confirmed cyber incidents in 2024 rule out “strong”.
What is documented
- TÜV Rheinland certified the development process at the Lake Forest site under IEC 62443-4-11
- Schneider is a CNA, publishes monthly security notifications and has a coordinated disclosure policy6
- CISA and CERT-FR picked up Schneider advisories with available fixes7 8
- According to the company, AI applications are tested on a risk basis, with shadow deployment for critical infrastructure3 4
The case against
Cyber incidents
- In a 2024 ransomware attack by the Cactus group, data of the Sustainability Business Division was exfiltrated; Resource Advisor was down until the end of January12 13
- In 2024 the Hellcat group gained access to an internal Jira system; Schneider confirmed the access, but not the data theft15 14
- In 2025 the Cl0p group named Schneider as a victim; no confirmation was found17
Not AI-specific
- The independent evidence covers AI systems only indirectly; independent tests or red teaming of the copilots are missing1
- Usage restrictions, misuse monitoring and reporting channels for AI incidents are not documented
Sources for Safety (17)
- TÜV Rheinland Industrie Service GmbH / ISASecure: ISASecure SDLA Certificate No. 968/SDLA 107.05/26 – Schneider Electric Systems USA, Inc., Lake Forest · 11 February 2026 · primary source
- Schneider Electric (Blog): SDL: Using a secure-by-design approach to build customer trust · 5 March 2025 · secondary source
- Schneider Electric (Blog/Podcast «AI at Scale»): Where does AI governance stop? – Ania Kaci: AI governance decoded · 21 April 2025 · secondary source
- Energy Digital (Energy Magazine, Juni 2026): Energy Magazine June 2026, p. 150 (interview with Philippe Rambach) · 1 June 2026 · secondary source
- Memoori Research: Schneider Electric AI Strategy in Smart Buildings: An Independent Assessment · 3 May 2026 · primary source
- Schneider Electric SE (Corporate Product CERT): Schneider Electric Vulnerability Handling and Coordinated Disclosure Policy V3.0 · 19 July 2022 · primary source
- Cybersecurity and Infrastructure Security Agency (CISA): CISA Releases Seven Industrial Control Systems Advisories · 18 March 2025 · primary source
- Leto (leto.legal): Schneider Electric: ANSSI warns of critical vulnerabilities in industrial equipment · 19 April 2026 · secondary source
- Schneider Electric SE: 2023 Sustainability Report · 3 November 2026 · secondary source
- Schneider Electric SE: 2023 Trust Report · 3 November 2026 · secondary source
- Schneider Electric SE: Integrated Report 2025 · 3 November 2026 · secondary source
- Schneider Electric SE: Sustainability Business Division of Schneider Electric Responds to Cybersecurity Incident · 29 January 2024 · secondary source
- BleepingComputer: Energy giant Schneider Electric hit by Cactus ransomware attack · 29 January 2024 · primary source
- The Register: Schneider Electric confirms cybersecurity incident (Hellcat) · 5 November 2024 · primary source
- SecurityWeek: Schneider Electric Launches Probe After Hackers Claim Theft of User Data · November 2024 · primary source
- Dark Reading: Schneider Electric Clawed by 'Hellcat' Ransomware Gang · 5 November 2024 · primary source
- OffSeq Threat Radar: Industrial Giants Schneider Electric and Emerson Named as Victims of Oracle Hack · 28 October 2025 · secondary source
3 · Transparency
Schneider Electric describes internal system cards with limits and data flows but does not publish them, and the announcements on the copilots contain no information on limits, validation or data use.
What is documented
The case against
Little disclosure on products
- System cards, model information and a transparency report are not published5
- The announcements on the copilots name the base model only in general terms, without limits, validation or data use1
- An independent analysis notes that efficiency claims of some AI building products have not been validated at scale6
- Rules on labelling AI-generated outputs towards users are not documented
Sources for Transparency (6)
- Schneider Electric SE: Schneider Electric drives Generative AI productivity and sustainability solutions by integrating Microsoft Azure OpenAI · 13 November 2023 · secondary source
- Schneider Electric SE: Schneider Electric announces multiyear initiative building an AI-native ecosystem for sustainability and energy management · 15 May 2025 · secondary source
- Machinery (machinery.co.uk): Schneider Electric unveils industrial Gen AI copilot in collaboration with Microsoft · 3 November 2026 · secondary source
- Schneider Electric (Blog, Nicolas Rambaud): Cyber governance and AI: Balancing opportunity and risk through trustworthy AI · 2 September 2025 · secondary source
- Schneider Electric (Blog/Podcast «AI at Scale»): Where does AI governance stop? – Ania Kaci: AI governance decoded · 21 April 2025 · secondary source
- Memoori Research: Schneider Electric AI Strategy in Smart Buildings: An Independent Assessment · 3 May 2026 · primary source
4 · Why the harm rule does not apply
The harm rule does not apply: the 2024 and 2025 cyber incidents are not AI-related and involve no judgment, authority decision or fine.
5 · Controversies (3)
- mediumThe Hellcat group gained access to an internal Jira system; Schneider confirmed the access, but not the alleged data theft. No link to AI. · 4 November 2024 · unklar
- mediumIn a ransomware attack by the Cactus group, data of the Sustainability Business Division was exfiltrated and Resource Advisor went down; a link to AI is not documented. · 17 January 2024 · unklar
- minorThe Cl0p group named Schneider Electric as a victim in connection with an Oracle compromise; no confirmation was found. No link to AI. · 28 October 2025 · unklar
6 · What remains open
- 2025 URD with corporate governance report, risk factors and litigation, which could not be retrieved
- AI mandate of the Digital Committee and the Audit & Risks Committee
- Published AI policy, risk classes and approval procedure for purchased AI
- Text of the agreement with the trade unions of 2026-05-22
- Forensic results and assessment by authorities of the 2024 cyber incidents
- Independent tests of the generative AI products
7 · How it was assessed
Cut-off date: 3 November 2026 · 40 searches, 41 fetches.
Primary documents read (11)
- Schneider Electric Integrated Report 2025 (full version, extended)
- Notice of filing of the 2025 URD (AMF/DILA)
- Schneider Electric 2023 Sustainability Report (up to approx. 100,000 characters)
- Schneider Electric 2023 Trust Report (up to approx. 100,000 characters)
- Vulnerability Handling and Coordinated Disclosure Policy V3.0
- ISASecure SDLA certificate from TÜV Rheinland (Lake Forest)
- Decision 24-D-09 of the Autorité de la concurrence
- CISA advisory dated 18 March 2025
- Schneider Electric statement on the 2024 Cactus incident
- industriAll document on the social partner agreement on AI (2026)
- Board committee page
Research of 3 November 2026 under Globalance AI Footprint rulebook version 0