Pre-IPO ESG and sustainability disclosure readiness for HKEX (AI issuer perspective)
A China-based AI or embodied-AI company preparing to list on HKEX needs, between filing and its first post-listing annual report, to bring the ESG and climate disclosure data chain, governance structure and external assurance to a publishable, diligence-ready state.
Prerequisites
- The listing board (Main Board or GEM) and the expected first reporting year are settled — disclosure intensity differs by board and by whether the issuer is a Hang Seng Composite LargeCap Index constituent.
- The baseline facts are established: the Exchange’s amended Listing Rules took effect on 1 January 2025; the New Climate Requirements phase in such that for LargeCap constituents scope 1 and scope 2 GHG disclosure is mandatory for financial years commencing on or after 1 Jan 2025, while other climate disclosures are “comply or explain” from FY2025 and mandatory from FY2026; other Main Board issuers are “comply or explain” from FY2025; GEM issuers are voluntary from FY2025.
- You know the ESG report must be published at the same time as the annual report — the hard scheduling constraint, from which everything is back-planned.
- If a US listing is also under consideration, be accurate about the status of the SEC climate disclosure rules (stayed, never effective, and subject to a rescission proposal) and do not prepare as if they were in force.
Stages (26–39 weeks, estimate)
1. Stage 1 — Choosing the disclosure basis and diagnosing gaps
3–5 weeksFix the basis before drafting. HKEX Appendix C2 has three operative parts: Part B mandatory disclosure requirements, Part C “comply or explain” provisions, and Part D climate-related disclosures; disclosure prepared in compliance with IFRS S1 and IFRS S2 is considered to comply with Part D, so preparing directly on the ISSB basis is often both simpler and more readable for international investors. The output is a gap diagnosis: which disclosures existing data can support, which metrics have no underlying data at all, and which require new processes. If a US listing is in scope, settle the status of the SEC climate rules here so effort is not spent on a false premise.
Obligations in this stage
- Track the actual status of the rule and do not prepare disclosure as if it were in forceSEC climate-related disclosure rules (stalled and proposed for rescission) · 89 FR 21668;89 FR 25804;91 FR 33296 · Severity Medium
Deliverables
Evidence to retain
- · Basis-selection memorandum comparing direct Appendix C2 preparation against preparation under IFRS S1/S2, with the conclusion
- · Applicability determination record (board, LargeCap constituent status, first reporting year)
- · Metric-by-metric gap diagnosis table (metric × existing data × gap × remediation × owner)
- · Multi-listing basis comparison, including a written conclusion on the status of the SEC rules
Common blockers
- · Treating the ESG report as a post-listing matter: the data chain needs a full financial year to produce comparable figures, and starting at filing is usually too late.
- · Copying a peer report structure: Appendix C2 Part B is mandatory, and borrowed structures routinely omit required elements.
- · Preparing against the US SEC climate rules: they never took effect and are the subject of a rescission proposal, so effort spent on them is wasted.
2. Stage 2 — Governance structure and materiality assessment
4–6 weeksAppendix C2 Part B opens with a board statement: the board’s oversight of ESG issues, the board’s ESG management approach, strategy and prioritisation, and how the board reviews progress against targets. For an AI issuer that statement must bring data and algorithm governance inside the ESG governance frame — including the appointment of a personal information protection officer and the periodic compliance audit arrangement — because investor diligence treats “data and algorithm governance” as the core ESG topic for such companies. In parallel, complete the materiality assessment and identify which climate-related risks and opportunities could reasonably be expected to affect cash flows, access to finance or cost of capital.
Obligations in this stage
- Appoint a personal information protection officer and run periodic compliance auditsPersonal Information Protection Law · 第五十二条、第五十四条;《个人信息保护合规审计管理办法》第四条、第十一条、第十二条 · Severity High
- Personal information protection impact assessment (PIPIA), retained for three yearsPersonal Information Protection Law · 第五十五条、第五十六条 · Severity Critical
Deliverables
Evidence to retain
- · Board ESG oversight terms of reference and meeting minutes
- · Materiality assessment workpapers (identification, prioritisation, conclusions and stakeholder engagement records)
- · Note bringing data and algorithm governance into the ESG topic set, covering the personal information protection officer and the compliance audit arrangement
- · ESG target-setting and progress-review mechanism documentation
Common blockers
- · A board statement ghost-written by ESG advisers with no actual board discussion: minutes that do not match the statement are a frequent finding in assurance and regulatory enquiry.
- · A materiality assessment that runs surveys without a financial link: Part D asks about reasonably expected effects on cash flows, access to finance or cost of capital.
- · An AI issuer writing generic manufacturing materiality topics: data governance, algorithmic fairness, content safety and compute energy use are this industry’s real topics.
3. Stage 3 — GHG inventory and the metrics data chain
6–10 weeksThe longest and most underestimated stretch of the whole exercise. Scope 1 and scope 2 are the mandatory starting point, and for an AI issuer scope 2 — purchased electricity, particularly for owned or rented compute — is usually where the emissions actually sit, requiring electricity invoices, rack power, PUE and cloud-provider disclosures to be joined into a checkable data chain. Scope 3 is frequently the largest category for companies running on heavy third-party compute, and even where it is not disclosed this period the data-acquisition path should be established now. The output must withstand third-party assurance tracing, so each figure needs its source, calculation method and reviewer recorded in the workpapers.
Obligations in this stage
- Retain greenhouse gas and climate data capability for other disclosure demandsSEC climate-related disclosure rules (stalled and proposed for rescission) · Severity Low
Evidence to retain
- · GHG inventory workpapers (scope 1 / 2 / 3, each line annotated with data source and calculation method)
- · Organisational and operational boundary definitions
- · Data-acquisition plan for compute-related electricity, covering both owned facilities and cloud providers
- · Emission factor selection record (source, version, applicable year)
- · Data quality control records (reviewer, review date, treatment of anomalies)
Common blockers
- · Cloud providers unable to supply electricity or emissions data attributable to your company: the most common data break for AI issuers, and one to raise as a contract clause at renewal.
- · No version record for emission factors: factor sets are updated annually, and missing versions make year-on-year figures incomparable and draw immediate assurance challenge.
- · Data scattered across finance, administration and IT on inconsistent bases: half the inventory effort comes from here, and the longer it waits the more it costs.
4. Stage 4 — AI-issuer-specific disclosures and the domestic compliance position
4–6 weeksTurn “our AI compliance position in China” into disclosable, diligence-ready material. Sponsors and investors check item by item: the security assessment and filing status for generative AI services, the algorithm filing number and its publication, implementation of labelling for AI-generated content, the data classification/grading and important-data determinations, and whether personal information protection impact assessments have been carried out. These are not only risk-factor content — they are the substance of the “data and algorithm governance” topic in the ESG report.
Obligations in this stage
- Security assessment and algorithm filing for services with public-opinion attributes or social-mobilisation capabilityInterim Measures for the Management of Generative AI Services · 第十七条;《互联网信息服务算法推荐管理规定》第二十四条、第二十六条;《人工智能生成合成内容标识办法》第十二条 · Severity Critical
- File the algorithm within ten working days, publish the filing number and run the security assessmentProvisions on the Administration of Algorithmic Recommendation in Internet Information Services · 第二十四条、第二十五条、第二十六条、第二十七条 · Severity Critical
- Submit labelling materials at filing and assessment, and conform to the mandatory national standardMeasures for Labelling AI-Generated and Synthetic Content · 第十一条、第十二条 · Severity High
- Build a classification-and-grading register and identify important dataData Security Law · 第二十一条;《促进和规范数据跨境流动规定》第二条 · Severity High
- Personal information protection impact assessment (PIPIA), retained for three yearsPersonal Information Protection Law · 第五十五条、第五十六条 · Severity Critical
Deliverables
Evidence to retain
- · Filing status list (service/algorithm × filing number × filing date × where published)
- · Security assessment reports and submission records
- · Implementation note and sampling results for AI-generated content labelling
- · Data classification and grading register and important-data determinations
- · Personal information protection impact assessment reports and their retention records
- · Record of domestic regulatory interactions (enquiries, remediation, outcomes)
Common blockers
- · Filing status inconsistent with what is actually live: shipping a new feature without the new-product security assessment is a recurring hard finding in diligence.
- · Withholding the filing number as a trade secret: the number is subject to publication requirements, and withholding it invites questions instead.
- · Compliance material scattered across product lines: listing diligence aggregates by legal entity, and a last-minute consolidation exposes inconsistent bases.
5. Stage 5 — Third-party assurance readiness and the investor diligence pack
4–7 weeksAppendix C2 allows issuers to seek independent assurance voluntarily to strengthen credibility, and requires the level, scope and processes adopted to be described clearly in the report. In practice institutional investors increasingly read the presence of assurance as a signal of data quality. Two jobs here: prepare traceable workpapers on a limited-assurance-first basis, and consolidate the governance, data and domestic-compliance threads into one investor ESG diligence pack.
Obligations in this stage
- Appoint a personal information protection officer and run periodic compliance auditsPersonal Information Protection Law · 第五十二条、第五十四条;《个人信息保护合规审计管理办法》第四条、第十一条、第十二条 · Severity High
Deliverables
Evidence to retain
- · Written agreement on assurance scope and level, including the intended assurance statement wording
- · Traceable workpaper pack in which every disclosed figure traces back to source evidence
- · Assurance provider’s preliminary enquiry list and the responses
- · Investor ESG diligence pack (governance, metrics, compliance position, targets and progress)
Common blockers
- · Defining the assurance scope too broadly: a first engagement should focus on the most traceable metrics such as scope 1 and scope 2, and over-broad scope invites a qualified conclusion.
- · Workpapers containing only summary tables and no source evidence: assurance providers trace on a sampling basis, and figures that cannot be traced are carved out of scope.
- · Vague wording on the assurance level in the report: Appendix C2 requires the level, scope and processes to be described clearly, and vagueness is itself a disclosure defect.
6. Stage 6 — Publishing the first report and establishing the annual cadence
3–5 weeksBack-plan against the hard constraint that the ESG report publishes at the same time as the annual report: audit and annual-report finalisation dates determine the ESG report’s finalisation date, and assurance work must complete before that. On publication, move immediately into the annual cadence: next year’s data collection starts at the beginning of the new financial year, not at annual-report season. Also stand up a mechanism for tracking changes in basis — both exchange rules and ISSB standards are amended, and multi-listed issuers must additionally track divergence between jurisdictions.
Obligations in this stage
- Track the actual status of the rule and do not prepare disclosure as if it were in forceSEC climate-related disclosure rules (stalled and proposed for rescission) · 89 FR 21668;89 FR 25804;91 FR 33296 · Severity Medium
Deliverables
Evidence to retain
- · Publication schedule (annual report finalisation × ESG finalisation × assurance completion × board approval dates)
- · The published ESG report and the contemporaneous annual report
- · Next year’s data collection plan and assignment of responsibilities
- · Tracker for changes in disclosure basis (exchange rules, ISSB standards, domestic exchange rules)
Common blockers
- · Scheduling the ESG report after the annual report: the rules require simultaneous publication, and publishing later is non-compliance.
- · Disbanding the team after publication: nobody maintains the data chain, year two is rebuilt from scratch at double the cost, and the comparatives are not comparable.
- · Being simultaneously listed onshore without aligning the two bases: the Shanghai/Shenzhen/Beijing exchange rules and the MOF basic standard have their own requirements, calling for one preparation exercise with two adaptations.
Common failure modes
- · Writing the ESG report in “compliance document” mode: it is first of all an investor-facing disclosure, where comparability, traceability and year-on-year consistency matter far more than polished prose.
- · Underestimating the data-chain build: the GHG inventory needs source evidence spanning a full financial year, and it is the critical path of the whole exercise.
- · Ignoring the simultaneous-publication constraint and scheduling ESG after the annual report.
- · An AI issuer copying a manufacturing topic list: data and algorithm governance, content safety and compute energy use are what investors in this sector actually press on.
- · Treating domestic filings and compliance status as an appendix to the risk-factor section: it is simultaneously the substance of the ESG governance topic and a focus of diligence.
- · Preparing against the US SEC climate disclosure rules: they are stayed, never took effect, and are the subject of a rescission proposal.
Sources (17)
- HKEX Main Board Listing Rules Appendix C2 — Environmental, Social and Governance Reporting Code (version effective 1 January 2025)Hong Kong Exchanges and Clearing Limited (HKEX) · Part D, paras 16–41 · published 2025-01-01 · retrieved 2026-09-12
- HKEX — “Exchange Publishes Conclusions on Climate Disclosure Requirements” news release (19 April 2024)Hong Kong Exchanges and Clearing Limited · published 2024-04-19 · retrieved 2026-09-12
- HKEX — Listing with HKEX (official listing applicant entry page)Hong Kong Exchanges and Clearing Limited · retrieved 2026-09-12
- IFRS S1 General Requirements for Disclosure of Sustainability-related Financial InformationInternational Sustainability Standards Board (ISSB) / IFRS Foundation · retrieved 2026-09-12
- IFRS S2 Climate-related DisclosuresInternational Sustainability Standards Board (ISSB) / IFRS Foundation · retrieved 2026-09-12
- SEC — The Enhancement and Standardization of Climate-Related Disclosures for Investors (final rule)U.S. Securities and Exchange Commission / Federal Register · published 2024-03-28 · retrieved 2026-09-12
- SEC — Climate-Related Disclosures: Delay of Effective DateU.S. SEC / Federal Register · published 2024-04-12 · retrieved 2026-09-12
- SEC — Rescission of Climate-Related Disclosure Rules (proposed rule)U.S. SEC / Federal Register · published 2026-06-03 · retrieved 2026-09-12
- Personal Information Protection Law of the PRC (full text)National People’s Congress of the PRC / npc.gov.cn · published 2021-08-20 · retrieved 2026-09-12
- Measures for the Administration of Personal Information Protection Compliance Audits (CAC Order No. 18)Cyberspace Administration of China · published 2025-02-14 · retrieved 2026-09-12
- Interim Measures for the Management of Generative AI Services (Order No. 15 of seven departments)Cyberspace Administration of China and six other departments · published 2023-07-13 · retrieved 2026-09-12
- Provisions on the Administration of Algorithmic Recommendation in Internet Information Services (Order No. 9)Cyberspace Administration of China, MIIT, Ministry of Public Security and SAMR · published 2022-01-04 · retrieved 2026-09-12
- Measures for Labelling AI-Generated and Synthetic ContentCyberspace Administration of China, MIIT, Ministry of Public Security and NRTA · published 2025-03-14 · retrieved 2026-09-12
- Provisions on the Administration of Deep Synthesis Internet Information Services (Order No. 12)Cyberspace Administration of China, MIIT and Ministry of Public Security · published 2022-12-11 · retrieved 2026-09-12
- GB 45438-2025 “Cybersecurity technology — Labeling method for content generated by artificial intelligence” (mandatory national standard)SAMR and Standardization Administration of China (national standards portal) · published 2025-02-28 · retrieved 2026-09-12
- Data Security Law of the PRC (full text)National People’s Congress of the PRC / npc.gov.cn · published 2021-06-10 · retrieved 2026-09-12
- Provisions on Promoting and Regulating Cross-border Data Flows (CAC Order No. 16)Cyberspace Administration of China · published 2024-03-22 · retrieved 2026-09-12
Content as of 2026-09-12 · corpus 2026.09|Compiled from public official sources with links and review dates for compliance decision support. Rules change; have licensed professionals review before external disclosure, regulatory filing or contractual commitments.