Have your say: Australian AI Training Scheme

Good Ancestors is consulting on a proposed Australian AI Training Scheme — a two-part proposal comprising an AI Training Permit and an Australian AI Public Benefit Fund.

We invite comment from creators, collecting societies and rights management organisations, AI companies, and the public. Responses are sought within three parameters:

Delivering Australian creator expectations for consent, control, and compensation.
Consistency with the Government's commitment not to create a TDM exception, while preparing for future copyright challenges emerging from AI.
Facilitating economically viable AI training in Australia, in the broader national interest.

All three parameters must be met simultaneously. This paper does not consider proposals that achieve one or two at the expense of the others.

Download PDF

Building AI while Compensating Creators

Consultation Paper on an Australian AI Training Scheme

Your response

About you

Full name *

Organisation

Email address *

Perspective *

Primarily from the perspective of creators or rights holdersPrimarily from the perspective of tech or AIOther

Attribution preference *

I'm happy for my comments to be attributed to me or my organisationI'm happy for my comments to inform GAP's policy work, but not to be attributed to me or my organisation publicly

Consultation questions

Prefer to respond in a document?

You can upload your submission as a PDF or Word file in the upload section below — instead of, or in addition to, answering the questions here. All questions below are optional.

Your answers below are saved automatically in this browser only, so you can come back and finish later. Nothing is sent to Good Ancestors until you press submit. Drafts expire after one month.

On the overall proposal

1. Does the proposed Australian AI Training Scheme — a permit and fund — represent a workable path to meeting all three parameters in the Scope section? If not, what parameter is not met and why?

2. Are there alternative arrangements that meet the three parameters? If so, what is that proposal, how does it meet the three parameters, and how does it compare to the proposed Australian AI Training Scheme?

On the AI Training Permit

3. The scheme proposes opt-out as the primary consent and control mechanism. What approach would best deliver an opt-out that is effective, practical for creators and enforceable by Government? If you prefer opt-in, how could that be achieved within the parameters?

4. Are there national-interest conditions that should be attached to the permit from day 1?

On the Australian AI Public Benefit Fund

5. Who should the fund support, and how should eligibility be defined? Should the fund focus on creators whose materials have been used to train AI, creators who have lost earnings because of AI, the ongoing viability of Australia’s creative industry, or other matters? Why?

6. What should the role of established organisations like collecting societies be in designing, distributing and managing remuneration through the fund to creators?

7. How should compensation under the fund be calculated? What approach best balances feasibility, fairness, and administrative cost?

Open question

8. Are there issues raised by this proposal that are not addressed in the paper but are relevant to meeting the three parameters set out in the Scope section?

Upload a document

Accepted formats: PDF, DOC, DOCX. Max 10MB.

Good Ancestors will use the information you provide to inform its policy work on AI and copyright. We will only attribute your comments publicly if you have selected the attribution option above. See our privacy policy for more detail.