Skip to content
GLOBALANCE INTELLIGENCEThe economy ahead.
This Week Index 28/100 Sign in
Money & Investments › GI 100 Index
The Great Value ShiftInvestment ThemesGI 100 IndexGI PortfolioStart Future Check
AI FOOTPRINTsolidASSESSMENT 2026GLOBALANCEINTELLIGENCE
Globalance AI Footprint

Assessment as of November 2026 · next assessment 2027.

Result of the assessment by the Globalance research team under the published methodology. Not a certification or quality label.

How we assess ›

How this rating is builtVideo · 0:52
GI 100 Index · Profile

Samsung Electronics

South KoreaEnablersAssignment FullData quality High

Key messagesolid AI Footprint

Samsung has set rules and responsibilities for artificial intelligence, says it tests its models before deployment and Seals images edited with artificial intelligence, but discloses hardly any test results or information on its models. Over the collection of viewing habits via smart TVs, Samsung reached an agreement with the state of Texas on a consent requirement, and further lawsuits on the matter are pending.

Disclosure: the research was supported by an AI model made by Anthropic. Samsung Electronics is connected to Anthropic (e.g. as investor, partner, supplier, customer or through people). The assessment was therefore reviewed by people.

Opportunities

  • Responsibilitysolid

    Samsung documents AI governance with an AI Governance Council (AI Strategy Team, Samsung Research, Compliance) escalating to an AI Strategy Committee chaired by the CEO, an AI Ethics Implementation Guide, mandatory training since 2024 and an ISO/IEC 42001 certification for the home appliances business (S06, S07, S12).

    Samsung documents AI governance with an AI Governance Council (AI Strategy Team, Samsung Research, Compliance) escalating to an AI Strategy Committee chaired by the CEO, an AI Ethics Implementation Guide, mandatory training since 2024 and an ISO/IEC 42001 certification for the home appliances business (S06, S07, S12). At board level there is a Sustainability Committee comprising all independent directors and an Audit Committee overseeing compliance, the whistleblowing system and data protection, but no explicit responsibility for AI (S15, S12). Significant limitations: Ranking Digital Rights found no evidence of due diligence or operational rules for algorithms (S33), Glass Lewis criticises the independence of the board (S16), and metrics, participation rates or external audits of AI governance are missing.

  • Safetysolid

    The AI Safety Framework describes safety assessments with benchmarks, four-phase red teaming in an isolated environment and re-assessments (S09), and information security is documented with a CISO, a monthly Information Security Committee and certifications such as ISO/IEC 27001, 27701 and SOC 2 (S26).

    The AI Safety Framework describes safety assessments with benchmarks, four-phase red teaming in an isolated environment and re-assessments (S09), and information security is documented with a CISO, a monthly Information Security Committee and certifications such as ISO/IEC 27001, 27701 and SOC 2 (S26). Independently documented are strong security and data breach disclosures and seven years of security updates (S33), as well as the research capability of Samsung Research in the DARPA competition (S28). Against this stand a weak external security rating (S16), data leaks at service providers (S24, S32) and medium-severity data protection proceedings over content recognition on smart TVs (S01, S36, S41), which rule out “strong”; test results, external red teaming and thresholds are missing.

Risks

  • Transparencylimited

    Samsung publishes AI principles and an AI Safety Framework and Seals AI-edited images with watermarks and C2PA metadata (S08, S09, S18).

    Samsung publishes AI principles and an AI Safety Framework and Seals AI-edited images with watermarks and C2PA metadata (S08, S09, S18). However, Ranking Digital Rights independently finds that rules on the use of algorithms and information on recommender systems are missing (S33). Model and data cards are mentioned only as internal instruments; for Gauss2 and Galaxy AI no public model cards, training data information or transparency reports were found (S07, S09, S21), and labelling applies only to individual features and is easy to remove (S18, S19).

  • Risk area

    Personal data in devices: smart TV viewing data and face templates

  • Risk area

    Generative features on smartphones: deception and harmful content

  • Risk area

    AI glasses with camera and microphone: protecting third parties from recording

  • No serious controversy recorded.

What lies behind the assessment

Research of 3 November 2026
42searches
41pages and documents read
12primary documents
6controversies checked
6open points, named openly
The full research: pillars, evidence, controversies and open points
AI Footprint · Enabler · As of 3 November 2026
Samsung ElectronicsOverall rating: solid

Samsung documents AI bodies up to a committee chaired by the CEO and a safety framework with red teaming. Test results, model cards and AI responsibility on the board are missing; lawsuits over data collection via smart TVs are pending.

Responsibility

40 %
limitedsolidstrong

AI bodies, guidance and mandatory training exist, but AI responsibility on the board is missing.

  • +AI Governance Council escalating to a committee chaired by the CEO Source ↗
  • +ISO/IEC 42001 certification for the home appliances business Source ↗
  • +AI ethics training mandatory since 2024 Source ↗
  • –No explicit AI responsibility on the board Source ↗
  • –Ranking Digital Rights: no operational rules for algorithms Source ↗
All sources for this pillar (10)

Safety

40 %
limitedsolidstrong

Red teaming and information security are described, but test results and external audits are missing.

  • +Four-phase red teaming and assessment of every model before deployment Source ↗
  • +CISO, Information Security Committee and ISO/IEC 27001 certification Source ↗
  • +Ranking Digital Rights rates security audits and breach process well Source ↗
  • –No test results, no external red teaming, no thresholds Source ↗
  • –Agreement with Texas on smart TV viewing data, further lawsuit pending Source ↗
All sources for this pillar (21)

Transparency

20 %
limitedsolidstrong

Principles and image labelling exist, but model cards and information on algorithms are missing.

  • +Watermarks and C2PA for generative image features from Galaxy S25 Source ↗
  • +AI principles and AI Safety Framework published Source ↗
  • –Ranking Digital Rights: no information on algorithms and recommendations Source ↗
  • –No public model cards for Gauss2 and Galaxy AI Source ↗
  • –Labelling only for individual features and easy to remove Source ↗
All sources for this pillar (11)

AI resource use

Whole group incl. data centres · reporting year 2025
Disclosure partialDetailsClose
33,509 GWh Electricity197.7 million m³ water26 % renewable117.5 million t CO₂e100% renewable electricity by 2050: 32.5%
Electricity
33,509 GWh+4 % on prior year
Water withdrawal
197.7 million m³
Renewable electricity
26 %
Total CO₂e
117.5 million t
Target
by
Status 2025
Company-wide net-zero emissions (scope 1+2)
2050
14.2 Mt CO₂e (prior year 14.9)
100% renewable power in device division and overseas sites
2027
Device division 94.8% (prior year 93.4%)
100% renewable power at all sites worldwide
2050
32.5% (prior year 31.4%)
Device division: replenish 100% of water consumed
2030
67,2 %

Targets and status as reported by the company. The chip division (about 98% of emissions, 26.2% renewable power) targets net zero only by 2050; for 2025 Samsung applied new Korean grid emission factors.

Samsung Electronics used 33,509 GWh of electricity group-wide in 2025 and withdrew 197.72 million m³ of water. The semiconductor division (DS) sourced only 26.2% of its electricity from renewables and emitted 13.81 million t CO2e (Scope 1+2).

  • 2026 Emissions rose two years running despite RE100 target, domestic 18.27 million t in 2025
Sources

Supplement without a score: not part of the Globalance AI Footprint.

What is missing for “strong”

These steps are in the company’s own hands.

1Responsibility
AI oversight on the board

Assign responsibility for AI risks to a board committee and have AI governance metrics externally audited.

2Safety
Publish test results

Publish the results of safety assessments and red teaming including thresholds and have them externally audited.

3Transparency
Publish model cards

Publish model cards for Gauss and Galaxy AI covering limits, data provenance and data flows to partners.

Context: 101 companies assessed

Distribution of overall ratings, no ranking
strong
none yet
solid
← Samsung Electronics is one of 35
limited

No company reaches “strong” today. Disclosure on how companies handle AI is still young across the market.

Controversies
6 2 medium, 4 minor

No deduction for controversies.

Largest risk area
Personal data in millions of devices

Smart TVs, the Gallery app and Galaxy AI process viewing, face and usage data, and proceedings over the collection of viewing data on smart TVs are pending.

How thoroughly assessed
42 Evidence

12 primary documents read, 42 searches.

Read the full reasoning and all sources
Samsung Electronics · Full reasoning

AI governance and safety framework documented, but few test results or model details

Overall rating “solid”. No deduction for controversies. The superscript numbers link directly to the source.

1 · Responsibility

solidWeight 40%

Samsung documents an AI Governance Council escalating to an AI Strategy Committee chaired by the CEO and an ISO/IEC 42001 certification for home appliances, but Ranking Digital Rights found no operational rules for algorithms.

What is documented
  • An AI Governance Council of the AI Strategy Team, Samsung Research and Compliance escalates to an AI Strategy Committee chaired by the CEO1 2
  • An AI Ethics Implementation Guide with self-assessment checklists and a diagnostic process for generative AI cover the life cycle1 2 7
  • The ISO/IEC 42001 certification applies to the home appliances business1
  • The Audit Committee oversees the Compliance Officer, the whistleblowing system and data protection; AI is named as a board competence area5
  • Compliance metrics are included in management remuneration3
The case against
Gaps at the top
  • Neither the board nor a committee is explicitly responsible for AI risks5 3
  • Glass Lewis criticises the independence of the board6
Implementation hardly documented
  • Ranking Digital Rights found no evidence of due diligence or operational rules for algorithms10
  • Training rates, AI metrics and an external audit of AI governance are missing; the AI Ethics chapter of the 2026 sustainability report remained unread3
  • Risk classes and approval procedures for models purchased from external providers are not described
Sources for Responsibility (10)
  1. Samsung Electronics: Samsung Electronics AI Ethics (PDF, English) · 3 November 2026 · primary source
  2. Samsung Electronics: Tech for All > AI Ethics: Implementing our AI Ethics Principles (web page, last updated 2026-06-26) · 26 June 2026 · primary source
  3. Samsung Electronics: Samsung Electronics Sustainability Report 2026 (English) · 3 November 2026 · primary source
  4. Samsung Electronics: Samsung Electronics Sustainability Report 2026 (Korean) · 3 November 2026 · primary source
  5. Samsung Electronics (via London Stock Exchange RNS): Corporate Governance Report FY2024 · 30 May 2025 · primary source
  6. Glass Lewis: Proxy Paper Samsung Electronics, Annual General Meeting 19 March 2025 · 25 February 2025 · primary source
  7. Samsung Electronics: Tech for All > AI Ethics: Samsung AI Ethics Principles (web page, last updated 2026-06-26) · 26 June 2026 · secondary source
  8. Samsung Electronics: Samsung Electronics AI Safety Framework (PDF, English) · 3 November 2026 · primary source
  9. Regulations.AI: Samsung AI Safety Framework (position overview) · 3 November 2026 · secondary source
  10. Ranking Digital Rights: 2025 Big Tech Edition: Samsung · 13 May 2025 · primary source

2 · Safety

solidWeight 40%

The AI Safety Framework describes safety assessments and red teaming, and information security is certified; a weak external security rating, data leaks at service providers and the proceedings over smart TV data rule out “strong”.

What is documented
  • Before every deployment the AI Safety Framework provides for benchmarks on toxicity, bias, truthfulness and reliability, adversarial testing and re-assessments3
  • Information security is documented with a CISO, a monthly Information Security Committee and certifications such as ISO/IEC 27001, 27701 and SOC 210
  • Ranking Digital Rights rates security audits, account security and the data breach process positively; devices receive seven years of security updates12 20
  • A team including Samsung Research won DARPA's AIxCC AI cybersecurity competition6
  • The BIPA class action over face recognition in the Gallery app was finally dismissed17
The case against
Testing not documented
  • Test results, external red teaming and risk thresholds are missing, although the Seoul commitments provide for them4 5
  • According to Glass Lewis, an external security rating places Samsung in the bottom tenth of the industry7
Data and privacy cases
  • Texas sued over the collection of viewing data via content recognition without consent; the agreement requires explicit consent13 14
  • A new class action on the same matter is pending in California; the allegations are unproven19 21
  • Customer data leaked in 2023 at Samsung UK via a third-party application and in 2025 at Samsung Germany via a service provider11 8
Sources for Safety (21)
  1. Samsung Electronics: Samsung Electronics AI Ethics (PDF, English) · 3 November 2026 · primary source
  2. Samsung Electronics: Tech for All > AI Ethics: Implementing our AI Ethics Principles (web page, last updated 2026-06-26) · 26 June 2026 · primary source
  3. Samsung Electronics: Samsung Electronics AI Safety Framework (PDF, English) · 3 November 2026 · primary source
  4. Regulations.AI: Samsung AI Safety Framework (position overview) · 3 November 2026 · secondary source
  5. UK Government (DSIT): Frontier AI Safety Commitments, AI Seoul Summit 2024 · 21 May 2024 · primary source
  6. DARPA: AIxCC results · 8 August 2025 · primary source
  7. Glass Lewis: Proxy Paper Samsung Electronics, Annual General Meeting 19 March 2025 · 25 February 2025 · primary source
  8. heise online: Data leak: 270,000 Samsung customer tickets on the darknet · 29 March 2025 · secondary source
  9. SecurityWeek: Samsung Bug Bounty Programme Payouts Reach $5m, Top Reward Increased to $1m · 3 November 2026 · secondary source
  10. Samsung Electronics: Tech for All > Security (web page, last updated 2026-06-26) · 26 June 2026 · primary source
  11. TechCrunch: Samsung says hackers accessed customer data during year-long breach · 16 November 2023 · secondary source
  12. Ranking Digital Rights: 2025 Big Tech Edition: Samsung · 13 May 2025 · primary source
  13. Office of the Texas Attorney General: Attorney General Paxton Secures Major Agreement with Samsung to Ensure Texans Are Protected from Smart TVs · 26 February 2026 · primary source
  14. Secure Privacy: Samsung, LG and the Texas ACR settlements · 26 February 2026 · secondary source
  15. Samsung Mobile Press: Your Privacy, Secured: How Galaxy AI Empowers You to Take Control of Your Data · 9 September 2025 · secondary source
  16. Newsis: “It could become spy glasses”: data protection commission asks Samsung for privacy safeguards for its AI glasses · 3 August 2026 · secondary source
  17. U.S. Court of Appeals for the Seventh Circuit (über FindLaw): G.T. et al. v. Samsung Electronics America, Inc. and Samsung Electronics Co., Ltd., No. 25-1120 (via FindLaw) · 7 August 2026 · primary source
  18. Tech-ish: Galaxy S26 AI ambition comes with a multi-cloud privacy cost · 2 March 2026 · secondary source
  19. Korea Daily (미주중앙일보): Class action over unauthorised collection of viewing data by Samsung smart TVs · 13 October 2026 · secondary source
  20. Samsung Electronics: Tech for All > Privacy: Data Breach Response Process (web page, last updated 2026-06-26) · 26 June 2026 · primary source
  21. Bloomberg Law: Samsung Customer Drops Suit Over Data Collection From Smart TVs · 15 July 2026 · secondary source

3 · Transparency

limitedWeight 20%

Samsung Seals AI-edited images with watermarks and C2PA, but Ranking Digital Rights found no rules on the use of algorithms, and there are no public model cards for Gauss2 and Galaxy AI.

What is documented
  • Samsung has published AI ethics principles and an AI Safety Framework and commits to explaining the purpose, functions and limits of AI4 10
  • Five generative image features from Galaxy S25 apply a visible watermark and C2PA metadata5 7
  • Model and data cards are described as internal documentation standards9 10
The case against
Missing disclosure
  • Ranking Digital Rights found no rules on the use of algorithms and no information on recommender systems8
  • No public model cards, training data information or transparency reports were found for Gauss2 and Galaxy AI11
Incomplete labelling
  • Labelling applies only to individual features and, according to critics, is easy to remove5 6
  • According to Texas, clear notices on content recognition in smart TVs were only required with the agreement1 2
Sources for Transparency (11)
  1. Office of the Texas Attorney General: Attorney General Paxton Secures Major Agreement with Samsung to Ensure Texans Are Protected from Smart TVs · 26 February 2026 · primary source
  2. Secure Privacy: Samsung, LG and the Texas ACR settlements · 26 February 2026 · secondary source
  3. Samsung Electronics: Samsung Electronics AI Ethics (PDF, English) · 3 November 2026 · primary source
  4. Samsung Electronics: Tech for All > AI Ethics: Samsung AI Ethics Principles (web page, last updated 2026-06-26) · 26 June 2026 · secondary source
  5. Android Authority: Samsung Galaxy S25 Content Credentials (C2PA) · 5 February 2025 · secondary source
  6. PetaPixel: Samsung's Image Authenticity Approach is Backward and It's AI's Fault · 13 February 2025 · secondary source
  7. Samsung Mobile Press: Your Privacy, Secured: How Galaxy AI Empowers You to Take Control of Your Data · 9 September 2025 · secondary source
  8. Ranking Digital Rights: 2025 Big Tech Edition: Samsung · 13 May 2025 · primary source
  9. Samsung Electronics: Tech for All > AI Ethics: Implementing our AI Ethics Principles (web page, last updated 2026-06-26) · 26 June 2026 · primary source
  10. Samsung Electronics: Samsung Electronics AI Safety Framework (PDF, English) · 3 November 2026 · primary source
  11. Asia Economy (asiae.co.kr): Smarter 'Gauss' Released... Samsung Electronics Unveils 2nd Generation Generative AI Model · 21 November 2024 · secondary source

4 · Why the harm rule does not apply

The harm rule does not apply: the agreement with Texas concerns content recognition on smart TVs, which the sources do not describe as AI, and establishes no violation. The index committee should review this classification.

5 · Controversies (6)

6 · What remains open

  • AI Ethics, Privacy Protection & Security and Compliance chapters of the 2026 sustainability report, which were truncated on retrieval
  • Sections on litigation and sanctions in the 2025 Korean business report
  • Board body responsible for AI risks and how often it deals with them
  • Scope of the AI Safety Framework and red teaming for purchased models and AI in the semiconductor business
  • Results of safety assessments, training rates and number of AI incidents
  • Status of the smart TV data class action in California and of further authority proceedings

7 · How it was assessed

Cut-off date: 3 November 2026 · 42 searches, 41 fetches.

Primary documents read (12)
  • Samsung Electronics Sustainability Report 2026, English (pp. 1–27, two retrievals)
  • 삼성전자 지속가능경영보고서 2026, Korean (pp. 1–18 incl. table of contents)
  • 2025 Business Report, English version of the 사업보고서 (pp. 1–40)
  • Corporate Governance Report FY2024 (complete, two retrievals)
  • Samsung Electronics AI Ethics (PDF) and AI Safety Framework (PDF)
  • Webpages: AI Ethics Principles, Implementing AI Ethics, Security, Privacy/Data Breach Response Process (as at 26 June 2026)
  • Seventh Circuit judgment: G.T. v. Samsung, No. 25-1120 (7 August 2026)
  • Press release by the Texas Attorney General on the Samsung settlement
  • Glass Lewis Proxy Paper AGM 2025
  • Ranking Digital Rights 2025 Big Tech Edition, Samsung company page
  • DARPA AIxCC results
  • UK Government: Frontier AI Safety Commitments

Research of 3 November 2026 under Globalance AI Footprint rulebook version 0

We assess documented practice in dealing with AI; this is not an investment recommendation. Three pillars, weights 40/40/20, harm rule.

Research of 3 November 2026 under Globalance AI Footprint rulebook version 0. Documented observations from public sources, no investment recommendation.

Important Notice. This material is provided for information purposes only and does not constitute an offer or solicitation to buy or sell any financial instrument, nor does it constitute investment advice or a recommendation. The GI 100 Index is a rules-based strategy index of Globalance Bank. Investments involve risk, including the loss of capital invested. The final product documents (including FinSA) shall prevail in all cases.

Your company? Information on inclusion in the GI 100 Index →

What comes next