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 FOOTPRINTlimitedASSESSMENT 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

Z.AI Co Ltd

ChinaInnovatorAssignment CoreData quality Very high

Key messagelimited AI Footprint

Z.ai discloses bodies for ethics and data security and publishes its language models together with technical reports, but had not published the safety framework it promised in 2024 by mid-2026. A US agency rates its latest open model as the strongest open model to date for cyberattacks, and since 2025 the US government has kept the company on an export blacklist because of military links, which Z.ai disputes.

Opportunities

  • Transparencysolid

    Z.ai publishes open weights and technical reports with information on the volume and categories of training data (23T and 28.5T tokens), and for GLM-4.5 a safety benchmark with an openly named weakness in unfairness and bias (S07, S08, S09).

    Z.ai publishes open weights and technical reports with information on the volume and categories of training data (23T and 28.5T tokens), and for GLM-4.5 a safety benchmark with an openly named weakness in unfairness and bias (S07, S08, S09). The FLI describes Z.ai as more transparent than its Chinese peers (S04). Disclosure on limits and safety is limited: the GLM-5.3 model card and the GLM-5 report contain no section on safety or limits, system prompts and behaviour specification are not disclosed (Information Sharing D), and labelling of AI content is documented only as a contractual rule (S09, S08, S04, S21). Hence solid, with reservations.

Risks

  • Responsibilitylimited

    The 2025 ESG report documents a structure: board responsibility with an ESG and strategy committee, an ethics review commission covering model development from planning to decommissioning, algorithm rules and a stated ISO/IEC 42001 certificate without a named certification body (S01).

    The 2025 ESG report documents a structure: board responsibility with an ESG and strategy committee, an ethics review commission covering model development from planning to decommissioning, algorithm rules and a stated ISO/IEC 42001 certificate without a named certification body (S01). Independent evidence, however, shows material weaknesses: Zhipu had not published the safety framework it committed to in Seoul in 2024 by June 2026 (S05, S03, S04), and the FLI grades Governance & Accountability F, with no clear governance structure and no whistleblowing channel (S04). In addition, the company remains on the US Entity List because of military use, which it disputes (S02, S25). Hence limited because of A-grade weaknesses; the prospectus and the corporate governance report remained unread.

  • Safetylimited

    Elements of implementation are documented: CAISI found GLM-5.2 robust against public agent hijacking attacks and more robust than other open models, the FLI acknowledges some incident-response infrastructure, and the ESG report describes encryption, access controls, red-team testing and emergency drills (S12, S04, S01).

    Elements of implementation are documented: CAISI found GLM-5.2 robust against public agent hijacking attacks and more robust than other open models, the FLI acknowledges some incident-response infrastructure, and the ESG report describes encryption, access controls, red-team testing and emergency drills (S12, S04, S01). Against this stand material weaknesses with A- and B-grade evidence: GLM-5.2 refused none of the agentic exploit tasks and answered sensitive biology questions more often than US models, GLM-5.3 is the most cyber-capable open model, and risk assessment receives an F from the FLI (S12, S11, S04). An official finding that the 智谱清言 app collected personal data beyond user consent concerns sub-criterion c (S22, S23). Hence limited.

  • Risk area

    Open weights of models with high cyber capabilities

  • Risk area

    State and military-related use of the models

  • No serious controversy recorded.

What lies behind the assessment

Research of 3 November 2026
39searches
37pages and documents read
10primary documents
3controversies checked
6open points, named openly
The full research: pillars, evidence, controversies and open points
AI Footprint · Innovator · As of 3 November 2026
Z.AI Co LtdOverall rating: limited

Z.ai documents bodies for ethics and data security and publishes its models together with technical reports. The safety framework it promised in 2024 is missing, and according to a US agency its latest open model is the most cyber-capable of its kind.

Responsibility

40 %
limitedsolidstrong

Ethics and strategy bodies exist, but the promised safety framework is missing.

  • +Board responsibility and ethics review commission documented in ESG report Source ↗
  • +Signatory of the 2024 Seoul Frontier AI Safety Commitments Source ↗
  • –Promised safety framework not published by June 2026 Source ↗
  • –FLI: no clear governance structure, no whistleblowing channel Source ↗
  • –On the US Entity List because of military links, disputed by the company Source ↗
All sources for this pillar (12)

Safety

40 %
limitedsolidstrong

Some safeguards are documented, but dangerous capabilities are barely limited.

  • +CAISI: GLM-5.2 withstood public agent hijacking attacks Source ↗
  • +FLI acknowledges some incident-response infrastructure Source ↗
  • –GLM-5.3 the most cyber-capable open model to date, according to CAISI Source ↗
  • –GLM-5.2 refused none of the agentic exploit tasks Source ↗
  • –Official finding: chat app collected personal data beyond consent Source ↗
All sources for this pillar (13)

Transparency

20 %
limitedsolidstrong

Open weights and technical reports, but little information on limits and safety.

  • +Open weights and technical reports with information on training data Source ↗
  • +FLI: more transparent than its Chinese peers Source ↗
  • +GLM-4.5 report openly names the weakness in unfairness and bias Source ↗
  • –GLM-5.3 model card without a section on safety and limits Source ↗
  • –AI labelling documented only as a contractual rule Source ↗
All sources for this pillar (6)
What is missing for “strong”

These steps are in the company’s own hands.

1Responsibility
Publish a safety framework

Z.ai should publish the framework with risk thresholds it promised in 2024 and set up a whistleblowing channel.

2Safety
Test dangerous capabilities before release

Z.ai should have cyber and bio evaluations reviewed externally before releasing open weights and publish the results.

3Transparency
Publish system cards with limits

Z.ai should publish a system card with safety evaluation, limits and failure modes for every new model.

Context: 101 companies assessed

Distribution of overall ratings, no ranking
strong
none yet
solid
limited
← Z.AI Co Ltd is one of 66

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

Controversies
3 2 medium, 1 minor

No deduction for controversies.

Largest risk area
Cyber-capable models with open weights

According to CAISI, GLM-5.3 is the most cyber-capable open model to date, and its weights have been freely available since August 2026.

How thoroughly assessed
31 Evidence

10 primary documents read, 39 searches.

Read the full reasoning and all sources
Z.AI Co Ltd · Full reasoning

Ethics bodies and open models, but no framework for dangerous capabilities

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

1 · Responsibility

limitedWeight 40%

The ESG report documents an ethics review commission and board responsibility, but independent reviewers find neither a published framework nor a whistleblowing channel.

What is documented
  • The 2025 ESG report assigns ESG responsibility to the board and names four committees, including an ESG and strategy committee1
  • An ethics review commission covers model development from planning to decommissioning; there are also rules on algorithm safety and model retirement1
  • The company states certificates under ISO/IEC 42001, 27001 and 27701 without naming the certification bodies1
  • In Seoul in 2024, Zhipu signed the Frontier AI Safety Commitments as the only Chinese firm5 11
  • Z.ai answered the survey for the FLI AI Safety Index 20262
The case against
Commitments and oversight
  • According to the FLI, the safety framework with risk thresholds committed to in Seoul had not been published by June 20265 4 2
  • The FLI grades Governance & Accountability F and finds no clear governance structure and no whistleblowing channel2
  • The composition and powers of the ethics review commission and the certification bodies are not disclosed1
Military links
  • Since January 2025 the US Department of Commerce has listed seven Zhipu entities on the Entity List for supporting China's military modernisation; Zhipu disputes this10 12
Sources for Responsibility (12)
  1. Knowledge Atlas Technology JSC (Zhipu): 北京智譜華章科技股份有限公司 2025年環境、社會及管治報告 (first ESG report, reporting year 2025) · 3 November 2026 · primary source
  2. Future of Life Institute: AI Safety Index – Summer 2026 (Digital, 14 Jul) · 14 July 2026 · primary source
  3. HKEX / Knowledge Atlas Technology JSC: Announcement – Appointment of Overall Coordinators (Rule 12.01C) · 19 December 2025 · primary source
  4. Future of Life Institute: AI Safety Index – Summer 2025 · 17 July 2025 · primary source
  5. UK Government (DSIT): Frontier AI Safety Commitments, AI Seoul Summit 2024 · 21 May 2024 · primary source
  6. NYU Shanghai RITS (Utku Ege Tuluk): GLM-5.3: Z.ai Scales Post-Training, Holds the Weights Back · 17 August 2026 · secondary source
  7. Mervin Praison (mer.vin): GLM-5.3 Ships Open Weights After Two Weeks as an API-Only Model · 28 August 2026 · secondary source
  8. 新京报 (Beijing News), 罗亦丹: 智谱GLM-5.3发布,涌现网络安全能力 · 14 August 2026 · secondary source
  9. Z.ai: Z.ai Terms of Use · 14 April 2026 · secondary source
  10. U.S. Department of Commerce, Bureau of Industry and Security (reproduced from the Federal Register via FR Tracker): Addition of Entities to and Revision of Entry on the Entity List, 90 FR 4617 (Doc. 2025-00704) · 16 January 2025 · primary source
  11. Carnegie Endowment for International Peace (Scott Singer): DeepSeek and Other Chinese Firms Converge with Western Companies on AI Promises · 28 January 2025 · secondary source
  12. The Star (Reuters, Joe Cash): Chinese AI-related firm Zhipu says strongly opposes inclusion in US export control entity list · 16 January 2025 · secondary source

2 · Safety

limitedWeight 40%

CAISI found GLM-5.2 robust against agent hijacking, but it refused no exploit task, and GLM-5.3 is the most cyber-capable open model.

What is documented
  • CAISI found no successful agent hijacking of GLM-5.2; the model was more robust than other open models but less robust than leading US models4
  • According to the ESG report, encryption, multi-factor authentication, red-team tests and emergency drills protect the systems; by its own account there were no data breaches in 20251
  • The FLI acknowledges some meaningful incident-response infrastructure2
  • According to reports, Z.ai held back the weights of GLM-5.3 for two weeks to complete safety evaluation and hardening8 9 7
  • The terms of use set out prohibited uses, suspension rights and a reporting channel11
The case against
Dangerous capabilities
  • CAISI rates GLM-5.3 as the most cyber-capable open-weight model to date; the weights have been public since 28 August 20266 8
  • GLM-5.2 refused none of the ten agentic exploit tasks and answered sensitive biology questions more often and in more detail than US models4
  • The FLI grades risk assessment F; Z.ai's own evaluations of dangerous capabilities are not published2
Data and traceability
  • A state testing body found that the 智谱清言 app (version 2.9.6) collected personal data beyond user consent; remediation is unclear12 13
  • The vulnerability register now refers only to Chinese state platforms, without individual details10
Sources for Safety (13)
  1. Knowledge Atlas Technology JSC (Zhipu): 北京智譜華章科技股份有限公司 2025年環境、社會及管治報告 (first ESG report, reporting year 2025) · 3 November 2026 · primary source
  2. Future of Life Institute: AI Safety Index – Summer 2026 (Digital, 14 Jul) · 14 July 2026 · primary source
  3. GLM-4.5 Team, Zhipu AI & Tsinghua University (arXiv): GLM-4.5: Agentic, Reasoning, and Coding (ARC) Foundation Models (arXiv:2508.06471v1) · 8 August 2025 · primary source
  4. NIST, Center for AI Standards and Innovation (CAISI): Assessment of Z.ai's GLM-5.2 · 8 July 2026 · primary source
  5. SPLX (Zscaler), Mateja Vuradin: GLM-4.5. Can It Pass the Enterprise AI Security Test Where Kimi K2 Failed? · 11 August 2025 · secondary source
  6. NIST, Center for AI Standards and Innovation (CAISI): CAISI's Assessment of Z.ai's GLM-5.3 Cyber Capabilities · 17 September 2026 · primary source
  7. NYU Shanghai RITS (Utku Ege Tuluk): GLM-5.3: Z.ai Scales Post-Training, Holds the Weights Back · 17 August 2026 · secondary source
  8. Mervin Praison (mer.vin): GLM-5.3 Ships Open Weights After Two Weeks as an API-Only Model · 28 August 2026 · secondary source
  9. 新京报 (Beijing News), 罗亦丹: 智谱GLM-5.3发布,涌现网络安全能力 · 14 August 2026 · secondary source
  10. Z.ai Security: Z.ai Security Disclosure Ledger (cvd.z.ai) · 3 November 2026 · secondary source
  11. Z.ai: Z.ai Terms of Use · 14 April 2026 · secondary source
  12. DoNews (杨亮): 公安部通报35款违法收集个人信息APP 智谱清言、台铃电动在列 · 20 May 2025 · secondary source
  13. IT之家 (远洋): 官方通报 35 款 App 违法违规收集个人信息:涉多款剪辑 App · 20 May 2025 · secondary source

3 · Transparency

solidWeight 20%

Z.ai discloses weights, training data volume and benchmarks, but its latest models are released without a section on safety and limits.

What is documented
  • The technical reports on GLM-4.5 and GLM-5 describe architecture, training, token counts and source categories such as web, code, books and papers3 4
  • The weights of the GLM models are published on Hugging Face, as are evaluation tools5 3
  • The GLM-4.5 report gives a SafetyBench result and discloses unfairness and bias as the weakest area3
  • The FLI describes Z.ai as more transparent than its Chinese peers6
  • The terms of use require visible labelling of AI-generated output and prohibit removing AI Seals1
The case against
Gaps on limits and safety
  • The GLM-5 report and the GLM-5.3 model card contain no section on safety, limits or usage restrictions4 5
  • System prompts and behaviour specification are not disclosed; the FLI grades Information Sharing D6
  • The technical implementation of labelling and the registration numbers of the services are not documented1 2
Sources for Transparency (6)
  1. Z.ai: Z.ai Terms of Use · 14 April 2026 · secondary source
  2. 国家互联网信息办公室 (CAC): 国家互联网信息办公室关于发布生成式人工智能服务已备案信息的公告 · 2 April 2024 · primary source
  3. GLM-4.5 Team, Zhipu AI & Tsinghua University (arXiv): GLM-4.5: Agentic, Reasoning, and Coding (ARC) Foundation Models (arXiv:2508.06471v1) · 8 August 2025 · primary source
  4. GLM-5 Team, Zhipu AI & Tsinghua University (arXiv): GLM-5: from Vibe Coding to Agentic Engineering (arXiv:2602.15763v2) · 24 February 2026 · primary source
  5. Z.ai (zai-org) auf Hugging Face: Model card zai-org/GLM-5.3 · 27 August 2026 · primary source
  6. Future of Life Institute: AI Safety Index – Summer 2026 (Digital, 14 Jul) · 14 July 2026 · primary source

4 · Why the harm rule does not apply

The harm rule does not apply: during the period there is no judgment, regulatory decision or fine concerning harm caused by the company's own AI systems. The Entity List is an export-control classification of the company, and the finding on the chat app concerns data collection.

5 · Controversies (3)

6 · What remains open

  • Prospectus of 30 December 2025 with the chapters on risk factors, regulation, litigation and governance
  • Annual report 2025 with the corporate governance report
  • Frontier safety framework under the Seoul commitment and release criteria for open weights
  • System card or safety report on GLM-5.3 and GLM-5.2 including the content of the safety evaluation
  • Original notice of the official finding on the 智谱清言 app and the remediation
  • Registrations of Zhipu services in the CAC lists of generative AI services

7 · How it was assessed

Cut-off date: 3 November 2026 · 39 searches, 37 fetches.

Primary documents read (10)
  • ESG Report 2025 by Knowledge Atlas Technology JSC (S01, two passes)
  • Federal Register 90 FR 4617, Entity List (S02, reproduction)
  • HKEX announcements dated 2025-12-19 and 2025-12-30 (S28, S29)
  • Technical reports GLM-4.5 and GLM-5 (S07, S08)
  • Model card GLM-5.3 (S09)
  • CAISI assessments GLM-5.2 and GLM-5.3 (S11, S12)
  • FLI AI Safety Index Summer 2025 and Summer 2026 (S03, S04)
  • Frontier AI Safety Commitments (S05)
  • Z.ai Terms of Use and Security Disclosure Ledger (S21, S20)
  • CAC announcement 2024-04-02 (S30)

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