Why I Have Written to South Australia’s AI Royal Commission

The questions I have put to the Commission, the Government and two of the commissioners, and why I am making that correspondence public

Many who have followed my writing on FCP will probably understand why South Australia’s new Royal Commission into Artificial Intelligence has caught my attention. In August, I explored the emerging relationship between the South Australian Government, OpenAI and the proposed Royal Commission in The Fox Writing the Rules. At that stage, the commissioners had not been appointed and the final Terms of Reference did not yet exist.

Much of my recent work has returned to the same underlying questions: who holds power, who checks it, how relationships between government, institutions and private interests are disclosed, and whether ordinary people can still see clearly how significant decisions are being made. That was central to The Check We Never Built, and it remains central here.

That does not mean every institutional overlap is evidence of wrongdoing, or that every connection points to coordination. As I argued in Who Owns the Conversation?, institutions can overlap and shape the environment in which decisions are made without requiring a hidden agreement between everyone involved. It means I believe important questions should be asked early, while structures are still being built, and answered against the public record rather than speculation.

South Australia now has a Royal Commission examining the opportunities and consequences of artificial intelligence. The final Letters Patent do not deal only with productivity, innovation and economic growth. They also require the Commission to consider AI in “human centric” terms, including human needs, rights, values and wellbeing, together with accountability, systemic trust, civics and the democratic process.

If a government is asking a Royal Commission to help shape how artificial intelligence will affect our workplaces, schools, public services and wider society, then I think it is reasonable to ask who is doing that work, what interests they bring with them, how those interests are disclosed and managed, and how concepts such as human dignity, rights and wellbeing will actually be weighed.

I am not suggesting that the three commissioners are unqualified. Quite the opposite. Dr Iain Ross AO, Kate Pounder and Professor Christopher Manning bring substantial experience across law, industrial relations, technology policy, artificial-intelligence research and investment.

Nor am I suggesting that expertise gained through government, industry or investment is itself a conflict. A Royal Commission of this kind could hardly function if prior experience were disqualifying. The questions are more basic.

Can the public see how relevant interests are being managed? Does the panel have, or will it deliberately seek, the disciplinary breadth required to give substance to the Commission’s own language about human needs, rights, values and wellbeing? And where will the infrastructure consequences of expanding AI use be examined if they sit outside this Royal Commission?

That final question also connects with Who Pays for the Cloud?, where I looked at the physical infrastructure beneath AI: data centres, electricity, water and the wider costs that can accompany rapid technological expansion. The Royal Commission now treats data-centre matters as outside its Terms of Reference, so I want to understand where those wider infrastructure consequences will instead be examined.

There are also a few factual matters I would rather ask the people concerned directly than continue relying upon secondary reporting. For that reason, I have now written separately to the Royal Commission, the Department of the Premier and Cabinet, Kate Pounder and Professor Christopher Manning. I am publishing the correspondence below for transparency.

That decision does not depend on whether anyone replies, nor does it depend on whether my longer article is ultimately published. Readers should be able to see the questions I actually put, rather than simply being told later that questions were asked.

If responses arrive, I will add them to the public record and consider them on their merits. If an answer corrects or narrows something I presently understand, then my subsequent writing should change accordingly. That is part of the exercise.

My Christian worldview is part of why I care about this process. If every person bears the image of God, then those exercising public power deserve to be treated fairly, but they should also be accountable for how that power is exercised. Truth should not be bent to fit the conclusion I expected to reach. Asking the questions openly, giving people a fair opportunity to answer, and being willing to change my view when the evidence changes are, for me, part of the same obligation.

1. Royal Commission into Artificial Intelligence

Subject: Request for comment — AI Royal Commission, commissioner interests and the meaning of “human-centric”

Dear Commission Secretariat,

My name is Mark Neugebauer. I am an independent South Australian writer and publisher at FCP Australia, where I write about questions of faith, culture, government accountability and institutional power.

I am currently preparing an article examining the Royal Commission into Artificial Intelligence, particularly the Letters Patent issued on 1 October and the three commissioners appointed to conduct the inquiry.

I want to be clear about the position from which I am approaching this. I am not alleging that Dr Ross, Ms Pounder or Professor Manning have acted improperly. The article says plainly that all three bring substantial and relevant expertise. Nor am I arguing that prior government, industry or investment experience should disqualify someone from serving on a Royal Commission. Expertise necessarily comes from experience. My concern is narrower.

The Letters Patent ask the Commission not only to consider productivity, innovation and economic growth, but also to ensure that technological innovation remains “human centric”, with appropriate safeguards and accountability. They define human-centric AI as AI that prioritises human needs, rights, values and wellbeing, and separately require consideration of systemic trust, civics and the democratic process.

Those are significant words, and I am interested in how the Commission intends to give them practical meaning. I would therefore be grateful if the Commission could respond to the following questions.

1. Commissioner interests and conflict management

Will the Commission publish declarations of relevant commissioner interests, together with any conflict-of-interest or recusal arrangements that apply?

I ask because other Australian inquiries, including the Disability Royal Commission, have published commissioner conflict declarations and recusals. Dr Ross also chairs the Net Zero Economy Authority, which operates under a published conflict-of-interest framework and maintains a public register of interests.

If the AI Royal Commission does not intend to publish comparable material, what process is being used to identify and manage actual, potential or perceived conflicts, and who is responsible for that process?

2. Specific interests

Without seeking private financial information, can the Commission confirm whether the following relationships have been disclosed and how any relevant conflicts will be managed?

Ms Pounder worked during 2025 as an Australian policy liaison for OpenAI. Can the Commission confirm the period of that engagement and whether any current financial, advisory, consulting or other professional relationship with OpenAI remains?

Professor Manning is a General Partner of AIX Ventures and also holds a technical role with Moonlake AI, an AIX-backed company. How will the Commission deal with matters affecting sectors in which AIX holds investments?

For context, I have found no evidence that an AIX portfolio company is involved in South Australia’s current Proof of Value AI trials, and I am not suggesting otherwise. My question is about transparency and the management of interests, not an allegation of commercial benefit.

3. The Commission’s human-centric mandate

Does the Commission regard the Letters Patent’s human-centric direction as requiring dedicated evidence from disciplines such as applied ethics, human-rights law, philosophy of technology, children’s rights, disability rights or civil liberties?

If so, how will that expertise be incorporated into the Commission’s deliberations? Will specialist advisers be appointed, issues papers commissioned, or particular evidence sought?

I recognise that commissioners can and should hear evidence from specialists without themselves representing every discipline or affected community. My question is whether the Commission has a deliberate method for ensuring that questions of human dignity, rights, agency and democratic trust receive the same depth of consideration as questions of productivity, economics and technological capability.

4. Data centres and infrastructure

The Commission’s Terms of Reference page states that matters relating to data centres are not part of its Terms of Reference. The Letters Patent themselves do not contain an express data-centre exclusion.

Could the Commission explain the basis for that published interpretation?

Does the Commission also regard the wider energy, water and electricity-grid consequences of AI adoption as outside its remit, or is there scope to consider those effects where they arise from the use and adoption of AI?

5. Public submissions

Does the Commission intend to publish public submissions, subject to the usual privacy and confidentiality considerations?

I would genuinely like the Commission’s account to appear in the article rather than simply my interpretation of the public record.

I will quote or summarise any response accurately and in context. I am also happy to attribute a response to the Commission rather than an individual officer if that is preferred.

For transparency, I intend to publish this correspondence as a public record of the questions I have put, regardless of whether a response is received or whether the article is ultimately published.

If possible, I would appreciate an attributable response by 12 noon ACDT, Friday 9 October 2026. If additional time is required, please let me know.

Thank you for considering the questions.

Mark Neugebauer

FCP Australia

2. Department of the Premier and Cabinet

Subject: Request for comment — AI Royal Commission appointments, scope and OpenAI MOU

Dear Media Unit,

My name is Mark Neugebauer. I am an independent South Australian writer and publisher at FCP Australia.

I am preparing an article examining the establishment of South Australia’s Royal Commission into Artificial Intelligence, the commissioners announced on 1 October and the final Letters Patent.

I approach the subject sceptically, but not from an anti-government or anti-technology position. My interest is principally one of transparency and accountability: when government is moving quickly through a significant technological transition, I believe the public benefits from being able to see how the people, processes and safeguards surrounding that transition have been chosen.

The article does not allege wrongdoing by any commissioner or government official, and it acknowledges that all three commissioners are highly qualified. I would like to include the Government’s account of several matters before publication.

1. Selection of commissioners

The Premier said in August that the commissioners would be appointed following an international search.

Who conducted that search?

Was there a written selection brief, skills matrix or set of criteria used to assess candidates?

In addition to expertise in economics, labour markets, technology and AI, did those criteria include disciplines such as applied ethics, human-rights law, philosophy of technology, civil liberties or democratic governance?

How was the shortlist developed, who recommended the final appointments, and can the selection criteria be made public?

I am not suggesting that every affected constituency should have had a representative commissioner. My question is about the disciplinary breadth the Government considered necessary for a Commission whose Letters Patent expressly refer to human needs, rights, values, wellbeing and democratic trust.

2. Conflict and interest screening

What conflict-of-interest screening was undertaken before the appointments were recommended?

Were Ms Pounder’s 2025 engagement as an Australian policy liaison for OpenAI and Professor Manning’s roles with AIX Ventures and Moonlake AI disclosed and considered during that process?

If so, what arrangements were made to manage any actual, potential or perceived conflicts?

3. The August infrastructure language

On 10 August, the Premier said publicly that the Royal Commission was concerned with the use of AI rather than the physical infrastructure underpinning it, and said he did not want the inquiry caught up debating data centres.

The Department’s announcement dated 11 August nevertheless listed as a proposed focus:

“opportunities related to AI related infrastructure including the interrelationship with energy transformation, water usage and associated impacts (such as electricity grid implications).”

That wording does not appear in the final Letters Patent, and the Commission now states that matters relating to data centres are outside its Terms of Reference.

Could the Department explain how the final scope was settled and why the 11 August infrastructure wording was not carried into the Letters Patent?

I am not suggesting the omission was designed to avoid scrutiny. I am asking for the Government’s explanation of the documentary sequence.

4. Where infrastructure will be examined

Which State process will examine the cumulative relationship between AI adoption and the physical infrastructure required to support it, including data centres, electricity demand and water?

The South Australian Data Centre Strategy refers to proposed Data Centre and AI Infrastructure legislation. When does the Government expect that legislation to be introduced, and will there be an opportunity for public consultation on an exposure draft?

5. OpenAI memorandum of understanding

Is the memorandum of understanding signed between the South Australian Government and OpenAI publicly available?

If it is not currently published, will the Government release it?

Does the MOU create any commitments, formal or informal, relevant to matters being examined by the Royal Commission, or is the Commission entirely independent of that relationship?

I would genuinely prefer the Government’s account to appear directly in the article rather than readers receiving only my interpretation of the public record.

I will quote or summarise any response accurately and in context and am happy to attribute the response to the Department or Government rather than a named officer if preferred.

For transparency, I intend to publish this correspondence as a public record of the questions I have put, regardless of whether a response is received or whether the article is ultimately published.

If possible, I would appreciate an attributable response by 12 noon ACDT, Friday 9 October 2026. If additional time is required, please let me know.

Thank you for considering the questions.

Mark Neugebauer

FCP Australia

3. Right of reply — Kate Pounder

Subject: For forwarding to Ms Kate Pounder — right of reply regarding FCP Australia article

Dear Commission Secretariat,

Would you please forward the following right-of-reply request to Ms Kate Pounder? I would be grateful if you could confirm that it has been passed on.

Dear Ms Pounder,

My name is Mark Neugebauer. I am a South Australian writer and publisher at FCP Australia.

I am preparing an article examining the recently established Royal Commission into Artificial Intelligence and, because the article discusses your professional background directly, I wanted to give you an opportunity to correct anything I may have misunderstood or add context before publication.

The article describes you as highly qualified and tries to represent your policy record fairly. In particular, it notes that your work with the Tech Council combined support for AI adoption and economic opportunity with support for proportionate regulation and stronger privacy protections.

It also records that Commonwealth and Queensland Government documents identify you during 2025 as an Australian policy liaison for OpenAI.

Some secondary reporting suggests that engagement concluded in December 2025 following a six-month contract, but I have not found a first-party statement from you or OpenAI confirming the dates. I do not want to turn reported information into established fact simply by repeating it.

Could you therefore confirm:

• when your engagement with OpenAI commenced and concluded; and

• whether you currently retain any financial, advisory, consulting or other professional relationship with OpenAI?

The article also notes that the Royal Commission’s published biography does not mention the OpenAI engagement. I do not characterise that omission as concealment. My broader argument is simply that a public register of relevant commissioner interests would help distinguish ordinary biographical information from relationships that may be relevant to public confidence in the inquiry.

If there is anything else in that description you believe is incomplete or unfair, I would welcome the opportunity to correct it.

I will quote or summarise any response accurately and in context.

For transparency, I intend to publish this request as a public record of the opportunity I have given you to respond, regardless of whether a response is received or whether the article is ultimately published.

If possible, I would appreciate a response by 12 noon ACDT, Friday 9 October 2026. If you require more time, please let me know.

Thank you for considering the request.

Mark Neugebauer

FCP Australia

4. Right of reply — Professor Christopher Manning

Subject: For forwarding to Professor Christopher Manning — right of reply regarding FCP Australia article

Dear Commission Secretariat,

Would you please forward the following right-of-reply request to Professor Christopher Manning? I would be grateful if you could confirm that it has been passed on.

Dear Professor Manning,

My name is Mark Neugebauer. I am a South Australian writer and publisher at FCP Australia.

I am preparing an article examining South Australia’s Royal Commission into Artificial Intelligence and wanted to give you an opportunity to correct or contextualise several passages concerning your professional roles before publication.

The article describes your technical standing in AI as substantial and treats your connection with Stanford’s Institute for Human-Centered Artificial Intelligence as an important counterweight to one of the questions I am asking about the Commission’s human-centric mandate.

It also records that you joined AIX Ventures as a General Partner while on leave from Stanford and that you hold a technical role with Moonlake AI, which is backed by AIX.

The article discusses AIX’s investments only in relation to transparency and conflict management. For example, AIX has invested in companies operating in healthcare AI, while South Australia is separately trialling clinical AI transcription. I have found no evidence that an AIX portfolio company is involved in those South Australian trials, and the article says so expressly.

I am therefore not alleging that you or AIX stand to benefit from a particular South Australian recommendation.

My question is narrower: how should a Royal Commission make visible and manage circumstances in which one of its commissioners also has investment interests across sectors potentially affected by AI policy?

I would be grateful if you could clarify two factual points:

• Did you personally participate in AIX Ventures’ investment in Moonlake AI, and if so, in what capacity?

• Is there anything in the description of your AIX Ventures or Moonlake roles above that you believe requires correction or further context?

Any broader comment on how you intend to manage potential intersections between your investment work and your role as Royal Commissioner would also be welcome.

I will quote or summarise any response accurately and in context.

For transparency, I intend to publish this request as a public record of the opportunity I have given you to respond, regardless of whether a response is received or whether the article is ultimately published.

If possible, I would appreciate a response by 12 noon ACDT, Friday 9 October 2026. If additional time is required, please let me know.

Thank you for considering the request.

Mark Neugebauer

FCP Australia

What happens now

I have asked for responses by 12 noon ACDT on Friday 9 October.

If responses are received, I will make the fact of those responses clear and update the public record accordingly. More importantly, I will consider what they actually say.

If an answer establishes that one of my concerns was misplaced, the concern should narrow. If it provides information I did not previously have, that information should be incorporated. If it raises further legitimate questions, those can be examined on their own evidence.

If no response is received, I will record that fact. I will not treat silence as proof of a conflict, improper motive or absence of a process.

There is an important distinction between an unanswered question and an answered allegation. They are not the same thing.

For readers who have followed FCP for some time, this will probably look familiar. Much of my writing has been an attempt to understand not merely the decisions governments make, but the structures through which those decisions become possible and the checks that remain available to ordinary citizens.

Sometimes scrutiny confirms a concern. Sometimes it weakens one. Sometimes it changes the question entirely.

That is why I keep returning to the same simple principle:

show the evidence, show the questions, distinguish what is known from what is not, and remain willing to change course when better information appears.

If I expect that standard from institutions, I should be prepared to apply it to my own work as well.

Thanks for reading

God bless

Mark.