Event: Shaping Australia's AI Future: the Joint Select Committee on AI

This is intended to act as background reading and question preparation; not a finding of wrongdoing or influence.

Core discipline: Co-appearance, shared subject matter, a fiscal-sponsorship arrangement, or contact with the same public institution does not by itself establish coordination, influence, shared funding, political affiliation, or control.

This appears to be a public, online pre-submission briefing convened by AI Safety Australia & New Zealand (AISANZ), bringing together five policy-facing voices from charitable AI-safety advocacy, faith-based social-justice advocacy, catastrophic-risk/national-security policy advocacy, and responsible-AI practice to discuss Australia’s live parliamentary AI inquiry. The Luma listing says the session is intended to explain the inquiry, offer multiple perspectives, and help participants consider individual submissions before 14 September 2026. It is not documented as a committee hearing, a committee-sponsored event, or a forum whose speakers represent Parliament.

The listed host is Michael Kerrison, AISANZ Executive Director; the listed speakers are Greg Sadler and Emily Grundy (Good Ancestors), Mark Zirnsak (Uniting Church), Devon Whittle (Global Shield), and Emily Low (Gradient Institute). The event listing had “More TBC!”, so this is a briefing on the announced lineup, not necessarily the final programme.

Event, organiser, and inquiry at a glance

AISANZ describes itself as a regional AI-safety community and field-building initiative focused on careers, research, connections, and opportunities, with a mission to build local capacity to mitigate catastrophic AI risks. AISANZ states that it is fiscally sponsored by Good Ancestors. This is a direct institutional and donation-pathway relationship.

Parliament records that the Joint Select Committee on Artificial Intelligence was appointed by House and Senate resolutions on 20 August 2026. Written submissions close 14 September 2026, and the final report is due by 30 November 2026. It is an inquiry and reporting body, not an AI regulator (Source).

No reviewed source establishes that AISANZ, any panel organisation, any listed speaker, or the event itself has a formal role in the Committee; has been commissioned by it; will be invited to give evidence; has lodged a submission; or funds the Committee. An individual submission is an opportunity to place material before an inquiry, not evidence of access or outcome.

Who is on the panel?

D: Documented Fact | O: Organisational Document | S: Self Reported | I: Interpretation | U: Unknown/Not established

Person	Documented role and organisational affiliation	Disclosed institutional / funding context	Relevant stated AI or policy work	Confidence and caution
Michael Kerrison Host; not listed as a panellist	[D] AISANZ names him Executive Director, leading strategy and operations; his public LinkedIn also identifies the position from October 2025. Source: AI Safety Australia & New Zealand	[D/O] AISANZ says it is fiscally sponsored by Good Ancestors. No AISANZ board, separate legal identity, total funding, named donor, grant amount, or event-specific budget was found in the reviewed material. Source: AISANZ	[O] AISANZ’s stated work is AI-safety field-building: careers, research, networks, education, and regional access. The event is a public explanation of the inquiry and invitation to consider individual submissions. Source: AISANZ	High for host/leadership and fiscal sponsorship. [U] No evidence that Kerrison represents the Committee or that AISANZ controls speakers’ positions.
Greg Sadler	[D] CEO and listed board member of Good Ancestors Policy; director of Effective Altruism Australia. He is also listed on the Luma event under Good Ancestors. Source: Good Ancestors Source: Effective Altruism Australia	[D] Good Ancestors’ 2025 ACNC statement reports AUD 1,468,888 total revenue, including AUD 1,441,578 donations/bequests and AUD 0 government grants, plus aggregate key-management-person compensation—not Sadler’s own pay. [O] Good Ancestors says it receives charitable grants and paid research work and no AI-company grants, sponsorship, or other funding. Source: Australian Charities and Not-for-profits Commission Source: Good Ancestors	[D/O] Good Ancestors says Sadler leads its AI policy work, including the Australia AI Policy White Paper and AI Legislation Stress Test. A 2024 parliamentary hearing schedule identifies him as Good Ancestors CEO and coordinator of Australians for AI Safety. Its published positions include managing current and emerging harms, AI-safety capability, international cooperation, and accountability. Source: Parliament of Australia	High for Good Ancestors/EAA roles and 2024 parliamentary appearance. His Effective Altruism connection is a documented charitable/intellectual affiliation, not party membership. Organisational finance is not personal income.
Emily Grundy	[D] Good Ancestors lists her as Senior Policy Analyst; an ASPI biography identifies her as a Good Ancestors policy officer. The Luma listing identifies her with Good Ancestors. Source: Good Ancestors Source: Australian Strategic Policy Institute	[S] Her LinkedIn says she received a 2023 Effective Altruism Infrastructure Fund grant to evaluate AI Safety Support; it is a project-specific self-report. [D/O] Good Ancestors’ aggregate finances and funding statements are organisational and do not identify her remuneration or salary funder.	[D] Named author of Good Ancestors submissions on general-purpose AI/AI agents and cyber strategy recommending an AI Act, expert regulator/Australian AI Safety Institute, transparency, safeguards, and developer/deployer accountability. MIT FutureTech lists her as a co-author of the AI Risk Repository, described as 1,725 risks from 74 frameworks. Source: MIT FutureTech	High for Good Ancestors role, submission authorship, and MIT publication. [U] No verified board/advisory appointment or political-party role. Claims sourced only to LinkedIn should be treated as self-reported. Do not confuse her with the UK demographer of the same name.
Mark Zirnsak	[D] Senior Social Justice Advocate, Uniting Church in Australia, Synod of Victoria and Tasmania. The Attorney-General’s Department identifies him as 2026 civil-society co-chair of Australia’s Fourth Open Government Forum. Source: Uniting Church Vic/Tas Source: Attorney-General’s Department	[D] His documented employment is with the Uniting Church. [U] The reviewed sources do not disclose personal salary, grants, sponsors, donors, or consulting income.	[D] A Synod submission signed in his advocacy role supports stronger safeguards against AI-generated child sexual-abuse material and non-consensual intimate imagery, including developer, platform, and search safeguards, training-data controls, and safety testing. The Attorney-General’s Department says he has advocated AI regulation; his wider work includes modern slavery, tax justice, online harms, public integrity, and democratic reform.	High for Church role, public-forum role, and the signed submission. A Christian, faith-based social-justice orientation is documented. [U] No party membership, candidacy, or partisan role was established. He is best understood as a public-interest/harms advocate, not a technical AI developer.
Devon Whittle	[D] Australia Director, Global Shield Australia, leading advocacy with government and stakeholders on global catastrophic risk, including AI, food-system resilience, and critical infrastructure. Source: Global Shield	[O] Global Shield describes itself as a US 501(c)(4) dependent on donor support and affiliated with a separate 501(c)(3), Global Shield Institute, with separate boards, staff, and finances. It discloses neither named donors nor amounts. [U] No Whittle-specific funding or remuneration was found.	[D] In a signed ASPI article and an AI Safety Forum presentation, he supports mandatory guardrails for high-risk/general-purpose AI, monitoring and incident reporting, accountability, provenance, security standards, assessment, and international coordination. He links advanced AI to cyber and CBRN/national-security risks. Source: ASPI Source: AI Safety Forum Australia	High for Global Shield role and stated policy views. An East Asia Forum biography also says he “currently works” as Special Counsel at Watson Farley & Williams; reviewed sources do not resolve whether this is stale, concurrent, or part-time. [U] No party or campaign role established.
Emily Low	[D] Senior AI Specialist, Gradient Institute, joining its practice team in June 2026. Her documented earlier background includes UWA System Health Lab / Industrial Ontology Foundry work and Social Ventures Australia. Source: Gradient Institute	[D] Gradient’s 2024 ACNC statement reports AUD 2,308,116 total revenue, including AUD 607,111 government revenue/grants and AUD 1,700,957 service revenue. It does not attribute any funding or remuneration to Low. [U] No personal grants, donors, sponsorships, or political donations found. Source: Australian Charities and Not-for-profits Commission	[O] Gradient says her work supports responsible AI adoption and development, AI governance, risk frameworks, and policy questions. No reviewed source provides a direct personal statement by Low on a particular AI Act, frontier risk, open-source policy, or compute-governance proposal. Source: Gradient Institute	High for Gradient employment and its role description. [U] No documented board/advisory or party role and no direct personal policy statement on the specific proposals likely to arise. Several people share this name; only the Gradient/UWA/SVA identity is covered.

Connection map in words

The clearest structural chain is: AISANZ -> fiscal sponsor Good Ancestors -> Good Ancestors personnel Greg Sadler and Emily Grundy. AISANZ names Good Ancestors as fiscal sponsor, and Good Ancestors directly employs/identifies Sadler and Grundy in the roles above. The Luma listing then brings those two together with Zirnsak, Whittle, and Low under their stated organisational affiliations.

Outside that chain, the evidence shows event-level co-appearance and overlapping policy ecosystems, rather than one organisation directing the others. Gradient Institute operated the separate Australian AI Safety Forum 2026, which AISANZ says was sponsored by the Department of Industry, Science and Resources (DISR) and involved a committee containing people from Timaeus, Good Ancestors, the University of Sydney, and CSIRO Data61. That is an institutional connection around a different forum, not evidence that DISR funds AISANZ, that Gradient funds Good Ancestors or Global Shield, or that any of those bodies funds or directs this panel.

How the panel relates to the Committee’s terms of reference

The Committee’s formal remit contains 13 specified matters in paragraphs (a)–(m), plus “any other related matters.” The mapping below is about subject-matter relevance, not an assertion that the speakers speak for the Committee or have taken a position on each item.

Committee terms	Evidence-supported relationship to the panel	Important limitation
(a)–(d) Productivity, distributed benefits, innovation, adoption and SME/family-business barriers	AISANZ describes careers, research, community and access-building; Gradient describes Low’s remit as responsible adoption, governance, risk frameworks, and policy. These intersect with adoption and capability questions. Source: AI Safety Australia & New Zealand Source: Gradient Institute	[U] The reviewed personal material does not establish a detailed panellist position on SME compliance burdens, local uptake metrics, regional distribution, or sectoral productivity.
(e)–(f) Sovereign capability, R&D, investment/talent, data sovereignty and Australian domain models	AISANZ’s field-building and talent focus, Grundy’s research and policy work, and the forum materials’ research and data-centre context bear on national capability and institutional capacity. Source: AI Safety Australia & New Zealand Source: Good Ancestors	[U] No reviewed evidence shows the panel has a unified position on data sovereignty, Australian training data, investment settings, or model-commercialisation policy.
(g) Workforce, skills, education and job quality	AISANZ explicitly works on career guidance, professional pathways, research access and community; Low’s applied-adoption work is relevant to workplace implementation. Source: AI Safety Australia & New Zealand Source: Gradient Institute	This shows relevance, not a documented position on displacement, industrial relations, job quality, or education-system reform.
(h) Commonwealth adoption, transparency and accountability	Low’s Gradient role concerns governance, risk and policy; Sadler, Grundy and Whittle publicly advocate institutional governance, accountability, or incident response; Zirnsak’s work includes public integrity and accountability. Source: Gradient Institute Source: Good Ancestors Source: Global Shield Source: Attorney-General’s Department	[U] No reviewed source establishes a panel-endorsed blueprint for Commonwealth procurement, automated-decision transparency, or audit practice.
(i) Adequacy of laws and regulatory frameworks	This is one of the strongest links. Good Ancestors submissions authored by Grundy call for an AI Act, expert regulator / Australian AI Safety Institute, transparency and accountability; Whittle calls for guardrails, reporting and legal clarity; Zirnsak supports stronger safeguards against specified AI-enabled sexual harms. Source: Good Ancestors Source: Global Shield Source: Uniting Church Vic/Tas	These are attributed organisational documents or individual statements, not a single agreed panel platform. Low has no identified direct view on a specific legislative proposal.
(j) Intellectual property and copyright	The Committee expressly includes copyright and training data. Zirnsak’s institutional submission mentions reasonable controls over training data in a child-safety context. Source: Uniting Church Vic/Tas Source: Parliament of Australia — Committee terms of reference	[U] The reviewed material does not identify a dedicated copyright/creative-sector speaker or a developed panel position on licensing, remuneration, or cultural content.
(k) Harms — fraud, scams, deepfakes, children and vulnerable Australians	Strong direct relevance: Zirnsak’s signed submission concerns deepfake sexual material, child sexual-abuse material and non-consensual intimate imagery. Good Ancestors’ published positions include scams, misinformation, bias and deepfakes; Whittle discusses scaled misuse and incidents. Source: Uniting Church Vic/Tas Source: Good Ancestors Source: Global Shield	Different risks require different evidence and remedies. Concern about catastrophic risk does not substitute for expertise in survivor support, consumer enforcement, or platform operations.
(l) National security, cyber security, foreign interference, supply chains and critical infrastructure	Strong direct relevance for Whittle through work on AI, CBRN risks, cyber, reporting and critical infrastructure, and for Grundy through cyber strategy, biosecurity, compute governance and crisis preparedness. Source: Global Shield Source: Good Ancestors	[U] Neither event participation nor past public service establishes current government authority, classified access, or an official security-agency position.
(m) Emerging AI capability, regulator / Australian AI Safety Institute and intelligence/security response	Strong direct relevance. Good Ancestors policy work and Grundy-authored submissions call for an Australian AI Safety Institute; Whittle advocates monitoring, reporting, assessment and scalable regulatory capacity; AISANZ’s mission focuses on catastrophic-risk field-building. Source: Good Ancestors Source: Global Shield Source: AI Safety Australia & New Zealand	[U] No reviewed source establishes that this event is endorsed by, funded by, or speaking for the Australian AI Safety Institute, the Committee, or intelligence/security bodies.

Documented gaps in representation and evidence

These are questions for coverage and transparency, not allegations.

• Announced-lineup coverage. The named speakers are primarily policy, civil-society, and nonprofit/public-interest professionals. On the announced evidence, there is no clearly identified representative speaking principally for frontier-model development, AI product deployment in a business, SMEs, organised labour, copyright/creative industries, disability and consumer advocacy, First Nations data governance, regional communities, or a regulator. “More TBC!” means this may change; the point is limited to the announced roster.

• Technical and operational balance. Grundy has AI-risk research experience and Low has technical/formal-methods background, but no reviewed source identifies a listed speaker as a current frontier-model engineer or commercial AI-system developer. That could leave implementation cost, feasibility, red-teaming practice, and deployment operations less directly represented.

• Committee-facing evidence. The reviewed sources show thematic relevance but no confirmed collective AISANZ submission, individual submission to this Committee, committee invitation, committee membership, or funding relationship. Any such claim should wait for a primary record.

• AISANZ governance and finance. Fiscal sponsorship by Good Ancestors is disclosed, but a public AISANZ board, constitution, separate legal identity, donors, total budget, grant schedule, and event-specific funding were not located in the reviewed sources. This is an evidence-disclosure gap, not evidence of improper governance or funding.

• Individual finance and affiliations. Institutional disclosures are uneven. Good Ancestors and Gradient publish ACNC information; Global Shield describes donor dependence but does not name donors or amounts. No reviewed source supports inferring individual salary, personal funding, or political allegiance from those institutional facts.

• Known source ambiguity. Whittle’s Global Shield role coexists with an unresolved East Asia Forum statement that he “currently” works at Watson Farley & Williams. The correct treatment is to disclose the ambiguity, not choose a narrative. Names also require care: the reports specifically distinguish this Emily Grundy and Emily Low from same-name individuals in other fields.