Tag: board governance

  • San Francisco Declaration, AI governance and the Board: who is overseeing the AI supply chain you depend on?

    What is the San Francisco Declaration

    On 15 November 2026, Australia and nine other countries signed the San Francisco Declaration on the Secure Foundations of Enduring Prosperity at the Pax Silica Founders Retreat in San Francisco. According to the Australian Government’s announcement, the declaration is a United States-led commitment to work together on the supply chains that underpin artificial intelligence: critical minerals, energy, semiconductors, advanced manufacturing, data centres, logistics, and the AI models themselves. The signatories are the United States, Australia, India, Israel, Japan, the Republic of Korea, Qatar, Singapore, the United Arab Emirates and the United Kingdom. [1][2]

    Two points are relevant for boards. First, this is a political declaration, not a regulation. On the text published so far, it creates no legal obligation for Australian organisations. Second, it is nonetheless relevant for Australian organisations. In our opinion – when governments start treating AI supply chains as economic security infrastructure, follow-through in areas like procurement policy, export controls, investment screening and vendor ecosystems becomes more likely.

    The direction is worth watching even though the timetable is unknown.

    What Boards need to understand by “AI supply chain assurance”

    Directors are generally familiar with supply chain risk for physical goods. AI supply chain assurance is the same discipline applied to a less visible stack. When your organisation deploys an AI system, you are not buying a single product. You are depending on a chain that typically includes:

    • Compute and hardware: the advanced chips (and the minerals, energy and manufacturing behind them) that train and run models.
    • Foundation models: the large models your vendors build on, often developed overseas and updated remotely, sometimes without notice.
    • Data: training and tuning data of varying provenance, including data scraped, licensed or synthesised by third parties.
    • Platforms and integrators: the cloud providers, software vendors and consultants who package AI into the product you actually use.

    Supply chain assurance is the discipline of understanding that chain, testing its points of weakness, and being able to demonstrate that this has been done: where did this model come from, what is in it, who can change it, and what happens to your organisation if it is withdrawn, compromised or regulated away? Cyber security agencies now recommend artefacts such as an AI Bill of Materials (AIBOM) or a Software Bill of Materials for AI: a documented inventory of the models, datasets, components and dependencies inside an AI system, so that when a failure occurs or a vulnerability is disclosed, you can answer “are we affected?” quickly rather than after weeks of investigation. [6][10]

    Why oversight of the AI supply chain is already a governance responsibility, not just a geopolitics story

    Boards of AI deployers need not wait for new legislation. Frameworks that already exist point boards in this direction. Most are voluntary guidance rather than legal obligation, and we note this explicitly below:

    • Australia’s Voluntary AI Safety Standard is voluntary, not legally binding. Its 10 guardrails apply to organisations throughout the AI supply chain. Guardrail 8 calls for transparency with other organisations across the chain about data, models and systems, and the standard includes procurement guidance so deployers can align suppliers through contracts. [4]
    • The National AI Centre’s Guidance for AI Adoption (also guidance, not law) advises organisations to identify, document and communicate accountability for shared responsibility across the AI supply chain – model developers, system developers and deployers – and to assess risks arising from other parties in the chain. [5]
    • ISO/IEC 42001 is the international AI management system standard. The standard’s Annex A covers third-party and customer relationships, including allocating responsibilities across the AI life cycle and managing suppliers in line with the organisation’s responsible AI approach. [7]
    • The EU AI Act: Under Article 25, a deployer (or distributor, importer or other third party) takes on the legal obligations of a high-risk system’s provider if it puts its own name or trademark on the system, substantially modifies it, or changes its purpose so that it becomes high-risk. This applies only to AI within the Act’s scope, including AI used in the EU, and Annex III high-risk obligations apply from 2 December 2027. [8][11] Australian Boards with European exposure should ask whether management knows where the organisation sits in the chain.
    • The AICD and UTS Human Technology Institute’s Director’s Guide to AI Governance frames boards as ultimately accountable for the adoption, implementation and outcomes of AI, including clear accountability mapping and sufficient board AI literacy. [9]

    The San Francisco Declaration raises the stakes on each of these. In our opinion – if allied governments coordinate on trusted vendors, critical minerals and sensitive technology, the vendor and model choices available to Australian organisations may increasingly be shaped by geopolitics as well as price and performance. A deployer’s supply chain position can change quickly through commercial failure, cyber incident, export controls, or a vendor exiting a market. [1][3]

    What Boards should ask of management

    These are the assurances a Board can reasonably seek from management, who are responsible for carrying them out. They work within frameworks your organisation may already use. They are framed for deployers: organisations that buy and use AI rather than build it.

    1. AI supply chain map

    The Board should ask for a register of AI systems in use providing details on: the foundation model behind it, the vendor and integrator, the cloud and compute it runs on, and the countries involved at each layer.

    You cannot assure what you have not mapped.

    This also surfaces “shadow AI”: tools used by staff that fall outside procurement oversight. [5][9]

    2. AI Bill of Materials for critical systems

    For the AI systems that matter the most, the Board should expect management to request an AIBOM or equivalent from the vendor: the models, datasets, components and dependencies inside the system, with version and source. G7 Cybersecurity Working Group has published minimum elements for an AI SBOM, so there is an emerging benchmark to point vendors at. [6][10]

    3. Extend vendor due diligence with AI-specific questions

    Generic supplier assessments do not adequately address AI risk. The Board should ask whether the vendor’s assessments cover training data sources and rights, model evaluation and bias testing, security posture, whether your data is retained for further training, the vendor’s own upstream dependencies, and incident and vulnerability notification practices. Management should prefer models and vendors with genuine transparency about these points. [6]

    4. AI system transparency and change control

    AI systems change over time, often without the deployer’s involvement: models are retrained, components swapped, terms updated. The Board can ask management to ensure contracts require notice of material changes, disclosure of compromised components or vulnerabilities, flow-down of transparency obligations to the vendor’s own suppliers, and audit or information rights. The Voluntary AI Safety Standard’s procurement guidance is a practical checklist for this. [4][6]

    5. Establish accountability across the AI supply-chain

    For each significant AI system, management should record which party – developer, vendor, integrator, your organisation – is responsible for monitoring performance, human oversight, incidents and updates. Shared responsibility that is not documented is responsibility that no party holds. [5][7]

    6. Manage concentration and exit risk

    For each critical AI dependency, the Board should ask: what is our exposure if this vendor, model or compute supply is disrupted, whether by commercial failure, cyber incident, export control or geopolitical action? Where the exposure is material, management should consider diversification, contractual exit rights, data portability and tested fallbacks. The declaration’s focus on “reducing excessive dependencies” is, in our view, equally sound practice at organisational level. [3]

    7. Verify provenance rather than assuming it

    Management should maintain a model registry of specific versions that have been verified, tested and approved for use, re-verify before adopting new versions, use integrity checks where available, and treat web-sourced models and content as untrusted by default. [6]

    8. Track the policy signal; the Board keeps watch. The San Francisco declaration is the beginning of a process, not the end. The Board can ask management to track the follow-through

    Australian procurement guidance, export control changes, investment screening, and any move from voluntary frameworks toward mandatory expectations. Whether and when that shift happens is uncertain; our judgment is that voluntary frameworks often become expected practice over time, but that is an assessment, not a legal prediction. [1][2][4]

    The bottom line for Boards

    AI supply chain assurance sounds technical. At Board level it is a familiar governance question in a new domain: do we know what we depend on, who controls it, and what happens if it fails or is taken away?

    It does confirm that the resilience and trustworthiness of AI supply chains is now a stated priority for the governments. Boards that ask this question now, and hold management to clear answers, will be better prepared for whatever follow-through comes, and better protected against failures that will not wait for regulation.

    General governance information, not legal advice. Sources checked at the time of publication and are subject to change.

    References

    ABC News, “Australia signs US-led AI supply chain agreement amid growing contest with China”, 17 November 2026. https://www.abc.net.au/news/2026-11-17/australia-signs-us-led-ai-supply-chain-agreement/107392252

    Australian Government Department of Industry, Science and Resources, “Australia signs San Francisco Declaration to secure technology supply chains”, 17 November 2026. https://www.industry.gov.au/news/australia-signs-san-francisco-declaration-secure-technology-supply-chains

    United States Department of State, “Pax Silica” (initiative overview and declaration text). https://www.state.gov/pax-silica

    Australian Government Department of Industry, Science and Resources, “Voluntary AI Safety Standard” (voluntary guardrails, including guardrail 8 on supply chain transparency, and procurement guidance). https://www.industry.gov.au/publications/voluntary-ai-safety-standard

    National AI Centre, “Guidance for AI adoption: implementation guidance” (sections 1.2 Supply chain accountabilities, 3.1.2, 4.3 Supply chain transparency). https://www.ai.gov.au/staying-safe-and-responsible/essential-ai-practices/guidance-ai-adoption-implementation-guidance

    New Zealand National Cyber Security Centre, “Artificial intelligence and machine learning: Supply chain risks and mitigations”. https://www.ncsc.govt.nz/assets/guidance/Documents/Artificial-intelligence-and-machine-learning-Supply-chain-risks-and-mitigations.pdf

    ISO/IEC 42001:2023, Information technology – Artificial intelligence – Management system (ISO, standard metadata: https://www.iso.org/standard/81230.html). The standard text is paywalled and was not reviewed directly; descriptions of its Annex A third-party and supplier controls rely on published secondary implementation guidance consulted on 18 November 2026: isomanaged.com, “ISO 42001 Annex A.10: Third-party and customer relationships” (https://isomanaged.com/knowledge-base/iso-standards/iso-42001/iso-42001-annex-a-controls/iso-42001-annex-a-10-third-party-and-customer-relationships) and compliance.manzill.com, “A.10.3 Suppliers” (https://compliance.manzill.com/iso-iec-42001-2023/a.10.3/third-party-and-customer-relationships/suppliers/).

    European Commission AI Act Service Desk, “Article 25: Responsibilities along the AI value chain”, Regulation (EU) 2024/1689. https://ai-act-service-desk.ec.europa.eu/en/ai-act/article-25

    AICD and UTS Human Technology Institute, “A Director’s Guide to AI Governance”, Version 2, June 2026. https://www.aicd.com.au/news-media/research-and-reports/a-directors-guide-to-ai-governance.html

    German Federal Office for Information Security (BSI) for the G7 Cybersecurity Working Group, “Software Bill of Materials for AI – minimum elements”. https://www.bsi.bund.de/SharedDocs/Downloads/EN/BSI/KI/SBOM-for-AI_minimum-elements.pdf

    European Commission, “AI Omnibus enters into force” (27 July 2026; high-risk obligations under Annex III apply from 2 December 2027, Annex I embedded systems from 2 August 2028). https://digital-strategy.ec.europa.eu/en/news/ai-omnibus-enters-force

  • Buying frontier AI: the assurance Boards should demand now

    Australia’s proposed frontier AI rules sharpen a question every Board should already be asking: can we explain why the AI we buy is fit for the authority, data and decisions we give it?

    Stronger duties for model developers would be welcome. They would not replace the buyer’s governance.

    The organisation deploying AI still needs evidence about its own use, clear accountability and a workable response when something goes wrong.

    Australia’s AI adoption guidance makes that responsibility explicit. [1, 5]

    What is changing, and what is not

    On 10 November 2026, the Australian Government announced plans for mandatory safety and security protections applying to frontier AI companies wishing to operate in Australia. Its six proposed obligations cover security controls throughout development and deployment; management of serious model risks; independent model evaluation; rapid incident notification to government; published safety and security frameworks; and responsiveness to updated standards and rules. [1]

    The Government proposes a targeted AI safety regulator, working with the Australian AI Safety Institute for technical expertise and oversight. These are announced protections, not a basis for claiming that suppliers are already certified under a new regime. The release says legislative detail remains subject to further work and consultation; it does not specify commencement dates, numerical scope thresholds or penalty amounts. [1]

    Existing law is a separate matter. Directors’ duties, privacy, consumer, anti-discrimination and sector-specific requirements may already apply to an organisation’s use of AI. Which obligations apply depends on the organisation and the activity, not simply on whether its supplier calls a model “frontier”. [2]

    For financial services, the warning is already concrete. In April 2026, APRA reported overreliance on vendor presentations, opaque supply chains and limited evidence of tested exit arrangements. APRA’s expectations apply to its regulated entities. Other Boards can use the observations as a practical challenge to management, not as a claim that APRA regulates every sector. [4]

    The privacy deadline that cannot wait

    From 10 December 2026, APP entities must include additional privacy-policy information where they arrange for a computer program to use personal information to make, or do something substantially and directly related to making, decisions reasonably expected to significantly affect an individual’s rights or interests. [3]

    The required information covers the kinds of personal information used and the kinds of relevant decisions made solely by the program or supported in that substantial and direct way. Human review does not automatically take a process outside scope. This is not a blanket disclosure rule for every use of AI, nor is it confined to frontier models. Management should identify relevant decision processes now and obtain advice on coverage and any applicable exceptions. [3]

    Six questions that turn procurement into governance

    The following are WeGovernAi’s recommended buying controls, drawing on official guidance. They are not a claim that every control is a universal legal requirement. Scale the evidence and safeguards to the consequences of failure. [5, 6]

    1. What are we approving, and who owns it?

    Require an AI register that records purpose, accountable executive, affected people, permitted data and the authority given to the system. Distinguish assistance from action. Ask management to set acceptance criteria and unacceptable uses before selecting a supplier, supported by an impact and risk assessment. The Board should set direction and risk appetite; management should own implementation and demonstrate that controls work. [4, 5]

    2. What evidence supports this use, beyond the demonstration?

    Identify the application vendor, underlying model developer and material hosting and subcontractor dependencies. Request documentation, known limitations, evaluation methods and relevant independent assurance. Then test the complete service against the organisation’s own tasks and acceptance criteria. A strong model benchmark or a supplier’s safety framework is not a substitute for testing the system the organisation will actually operate. Record gaps and the residual-risk decision. [4, 5]

    3. What happens to our data, and what can we enforce?

    Map inputs, retrieved information, outputs and logs, including who can access them and where they are handled. Assess privacy, confidentiality, retention, deletion and training uses. Seek contract terms covering those matters, material model changes, assurance access, incident cooperation and exit assistance. “Not used for training” does not settle all data risks. OAIC recommends, as best practice, not entering personal information, especially sensitive information, into publicly available generative AI tools. Enterprise products still require due diligence. [4, 5, 6]

    4. Would we hear about an incident in time to act?

    Agree customer-notification triggers, timeframes and a monitored contact route. Connect supplier reporting to internal cyber, privacy, legal and operational response. The proposed frontier duties refer to notification to the Australian Government; buyers should not assume that this guarantees notice to them or fulfils their own reporting obligations. Rehearse suspension, investigation, evidence preservation and escalation before a critical deployment. [1, 5, 6]

    5. Can people intervene, and can the business continue?

    Give human overseers the knowledge, time and authority to challenge or stop unsafe operation. For agents, restrict permissions, log actions and require approval for high-impact changes. Test how access can be revoked. For critical processes, demonstrate a credible fallback and assess whether another provider is genuinely substitutable. Several applications can share the same underlying model dependency, so supplier counts alone do not establish resilience. [4, 5]

    6. What will the Board see after launch?

    Require reporting on material AI uses, performance, incidents, complaints, near misses, control failures and accepted risks. Retest material changes and review assurance independently of vendor marketing. A useful Board report should show what has changed, where limits were breached and what management did, rather than simply list tools deployed. Keep decisions, testing and approvals traceable throughout the service lifecycle. [4, 5]

    Use frameworks to organise evidence, not replace it

    Australia’s October 2025 Guidance for AI Adoption evolves and simplifies the 2024 Voluntary AI Safety Standard into six essential practices. The older ten guardrails remain a reference, but should not be presented as the only current Australian guidance or as enacted law. [7]

    Map one set of operational evidence to relevant frameworks: ISO/IEC 42001:2023 for an AI management system; NIST’s AI RMF for Govern, Map, Measure and Manage; and the OECD AI Principles for human rights, transparency, safety and accountability. These serve different purposes. A management-system certificate should not be treated as proof that a particular application is safe for the buyer’s intended use. [8, 9]

    A practical 90-day agenda

    As a management plan, not a regulatory timetable, WeGovernAi recommends:

    • First 30 days: identify consequential AI uses and hidden dependencies, name accountable owners and prioritise evidence gaps. Address the 10 December privacy-policy deadline before it arrives, not at the end of this period. [3, 5]

    • Days 31-60: assess critical suppliers, negotiate priority contract protections and test incident contacts and intervention controls. [4, 5]

    • Days 61-90: exercise a critical-process fallback, obtain independent challenge of key assurance gaps and bring residual-risk decisions to the appropriate governance body. [4, 5]

    The Board’s test is simple: can management show why the service is acceptable for its intended use, who owns the risk and how the organisation would contain a failure?

    If the answer depends entirely on the supplier’s confidence, the governance work is unfinished. Good procurement gives an organisation room to innovate without giving away control.

    General governance information, not legal advice. Sources checked on 15 November 2026.

    Sources and endnotes

    [1] Prime Minister of Australia. New Australian AI safety and security protections
    10 November 2026. Six proposed duties, proposed regulator and legislative status.
    https://www.pm.gov.au/media/new-australian-ai-safety-and-security-protections

    [2] Department of Industry, Science and Resources. Guidance for AI Adoption: The legal landscape
    Live guidance. Existing Australian laws that may apply to AI and supply-chain relationships.
    https://www.industry.gov.au/publications/guidance-for-ai-adoption/legal-landscape

    [3] Office of the Australian Information Commissioner. APP guidelines, Chapter 1
    Updated 30 September 2026. Paragraphs 1.35-1.52: 10 December commencement, scope, disclosures and human review.
    https://www.oaic.gov.au/privacy/australian-privacy-principles/australian-privacy-principles-guidelines/chapter-1-app-1-open-and-transparent-management-of-personal-information

    [4] Australian Prudential Regulation Authority. APRA Letter to Industry on Artificial Intelligence (AI)
    30 April 2026. Board oversight, supplier concentration, assurance and tested exit arrangements. APRA-regulated entities only.
    https://www.apra.gov.au/news-and-publications/apra-letter-industry-artificial-intelligence-ai

    [5] National AI Centre. Guidance for AI adoption: implementation guidance
    Live guidance. Practices 1-6, including accountability, impact assessment, supply chains, incidents, testing and human control. Buying controls adapt this guidance.
    https://www.ai.gov.au/staying-safe-and-responsible/essential-ai-practices/guidance-ai-adoption-implementation-guidance

    [6] Office of the Australian Information Commissioner. Guidance on privacy and the use of commercially available AI products
    Published 21 October 2024; page also displays an update of 17 January 2025. Product due diligence, privacy by design, ongoing review and the public-tool recommendation.
    https://www.oaic.gov.au/privacy/privacy-guidance-for-organisations-and-government-agencies/guidance-on-privacy-and-the-use-of-commercially-available-ai-products

    [7] Department of Industry, Science and Resources. Voluntary AI Safety Standard
    5 September 2024; live page checked. Ten guardrails and notice that the 21 October 2025 guidance evolves the standard into six practices.
    https://www.industry.gov.au/publications/voluntary-ai-safety-standard

    [8] National Institute of Standards and Technology. AI RMF Core
    AI RMF 1.0. Govern, Map, Measure and Manage; voluntary and context-based risk management.
    https://airc.nist.gov/airmf-resources/airmf/5-sec-core/

    [9] Organisation for Economic Co-operation and Development. AI principles
    Adopted 2019; updated 2024. Human rights, transparency, safety, accountability and lifecycle risk management.
    https://www.oecd.org/en/topics/ai-principles.html