Submission 197
Local & Independent News Association (LINA)
Submitter group: Creative sector.
An AI model identified the submitter group and wrote the summary, key points and recommendation summaries. Counts do not measure importance or merit.
Summary
Read the original submission for the full text.
Represents local and independent news publishers. Says existing copyright laws are broadly adequate but enforcement pathways are inaccessible to small newsrooms, and supports a legislative backstop for collective licensing, no text and data mining exemption, regulator-enforced training-data transparency and separate remuneration for training and output use. Describes operational benefits of AI alongside risks to journalism.
Key points
Quotes are the submitter’s own words. Text taken from documents, including scanned pages read by text-recognition software, can differ from the PDF. Check the original. Listed in the same order as on the submission’s page in the interactive guide.
Says copyright law gives members control in theory, but small newsrooms lack an accessible enforcement pathway.
In theory, copyright law gives LINA members the right to maintain control of their work and receive compensation for its use, but there is no accessible enforcement pathway for a small newsroom to actually use it, leaving a gap in how the mechanism operates in-practice.
Says voluntary licensing agreements should remain available but cannot be the only route to payment.
Voluntary agreements must remain available but cannot be the only route to payment.
Says LINA supports the Government's decision not to introduce a text and data mining exemption.
LINA also very strongly supports the Government's decision not to introduce a text and data mining exemption, and its position that Australian news should not be used to build or train AI without the control of the journalists and publishers who made it.
A program corrected obvious text-recognition errors in this quote for display.
Says collective and statutory licensing must be built to support small publishers, not just the largest.
Collective and statutory licensing must be built to support small publishers, not just the largest.
Says developers, rather than copyright holders, must disclose training use in enough detail for publishers to identify their work.
The resourcing and technical burden of this should not be placed on the copyright holder; rather, AI developers must be required to disclose use of copyright material in training, at a level of detail sufficient for a publisher to determine whether its work is included.
A program corrected obvious text-recognition errors in this quote for display.
Says AI can aid news gathering and operations, while requiring human oversight throughout publishing.
While we maintain that all content requires human intervention and oversight at each stage of the publishing process, AI tools can be useful in generating alerts on certain actions and announcements, compiling listings information, comparing large data sets and other mechanisms that can inform news gathering and news business operations.
A program corrected obvious text-recognition errors in this quote for display.
Recommendations
Recommendations are shown beside the submission’s own words. Listed in the same order as on its page in the interactive guide.
Ensure AI regulation of news keeps people central and addresses both fair intellectual-property remuneration and journalism's community-building role.
It is essential that any regulatory framework dealing with AI's impact on the news industry reinforces the importance of keeping people at the centre of journalism. It must deal with both the economic imperative to ensure that news companies are fairly remunerated for the use of their intellectual property, and the social imperative to support the community-building process of local journalism.
A program corrected obvious text-recognition errors in this quote for display.
Keep voluntary agreements available, alongside a genuine low-cost copyright enforcement pathway for local and independent news organisations.
Voluntary agreements must remain available but cannot be the only route to payment. Local and independent news organisations need a genuine, low-cost way to enforce their copyright, or existing laws and regulatory frameworks will continue to have little practical meaning for the majority of the sector.
Create a legislative backstop that lets small news organisations exercise their copyright, license their content if desired, and with greater oversight and regulation to ensure copyright compliance in the AI context.
LINA therefore supports the efforts of the Attorney General's office to create a legislative backstop that ensures that small news organisations can exercise their rights under copyright law, license their content if desired, and that there is greater oversight and regulation to ensure compliance to copyright in the AI context.
A program corrected obvious text-recognition errors in this quote for display.
Do not exempt freely accessible online news content or penalise publishers for making public-interest news accessible.
Similarly, no exemption should be made for news content that is freely accessible online. News produced by local and independent publishers is, by nature, bespoke. Journalists are covering stories that are not found in other sources. Further, those stories are in the public interest and are, by and large, freely accessible to communities as an essential information service. Putting genuine news content behind paywalls is often not in the public interest, particularly during emergencies, and policy-makers must take care not to punish news publishers for their accessibility. This news has inherent value to AI platforms, often as a single reference source on specific geographic and cultural topics.
A program corrected obvious text-recognition errors in this quote for display.
Treat news articles as archival material comparable to books and academic publications in their contribution to history.
In this way, news articles should be considered archival material comparative to books and academic publications in their contribution to history.
Support publishers both to bargain over training and retrieval licences if they choose and to opt out, particularly for cultural reasons.
As stated above, news organisations should be supported to bargain with AI companies to license content for training and retrieval purposes if they choose to do so, but they should also be supported if they elect to opt-out of providing that access, particularly for cultural reasons.
Avoid other exceptions, for example for 'research' purposes, that could have the same outcome as a text and data mining exemption.
The Government should be careful to avoid other exceptions, for example for “research” purposes, that could have the same outcome, transferring the entire value of Australian public interest journalism to model developers at no cost.
Address the use in Australia of AI models developed overseas using unlicensed Australian content, so that publishers and creators are compensated.
Finally, the Government needs to address the use in Australia of AI models developed overseas using unlicensed Australian content. Australian publishers and other content creators should receive compensation for this use of their work.
Build collective and statutory licensing to support small publishers, with payment distribution not based solely on scale (volume, traffic or measured usage).
Collective and statutory licensing must be built to support small publishers, not just the largest. This includes ensuring that any distribution of payments under a collectively licensed scheme must not be built solely on scale, including volume, traffic or measured usage.
Set a minimum payment per eligible newsroom under any collective licensing scheme, at a threshold that justifies the investment for small independent and local newsrooms.
Similarly, there should be a minimum payment per eligible newsroom. Engaging with copyright licensing will likely require an investment of labour and technical resources, regardless of the size of the news organisation. Therefore, minimum payments should be set at a threshold that justifies this investment for small independent and local newsrooms.
Separately remunerate use of news content for AI output, with attribution and working links as licence conditions.
As with use of news content for training AI models, use of content for AI output should be separately remunerated, with attribution and working links required as license conditions.
A program corrected obvious text-recognition errors in this quote for display.
Legislate transparency for both AI training and output use of news content.
For both purposes, transparency needs to be addressed in legislation.
Require developers to disclose training use of copyright material in enough detail for publishers to identify their work, supervised and enforceable by a regulator.
AI developers must be required to disclose use of copyright material in training, at a level of detail sufficient for a publisher to determine whether its work is included. This obligation should be supervised and enforceable by a regulator.
A program corrected obvious text-recognition errors in this quote for display.
Require AI companies to show their processes for adhering to data protections such as Robots.txt and at least the aggregated value of compensation paid to rightsholders in Australia.
AI companies should be required to show their processes for adhering to data protections such as Robots.txt and at least the aggregated value of compensation paid to rightsholders in Australia for the training and use of content.
A program corrected obvious text-recognition errors in this quote for display.
Do not extend copyright protection to material generated without a human author who contributed 'independent intellectual effort'.
Copyright protections should not be extended to material generated without a human author who contributed ‘independent intellectual effort’, as outlined in current copyright law.
Consider labelling requirements, including that editorial content produced with AI assistance be labelled.
it may be appropriate to introduce labelling requirements. Specifically to the news industry, LINA believes that it is appropriate that editorial content produced with the assistance of AI be labelled, and that content produced exclusively by AI is not journalism at all.
A program corrected obvious text-recognition errors in this quote for display.
Provide meaningful and technically effective mechanisms for rights holders to opt out of the use of their content, alongside a clear licensing and remuneration pathway covering both training and output use.
Where AI companies use copyright-protected Australian media content, there should be a clear pathway for licensing and remuneration. This needs to address both training of AI and output/Retrieval Augmented Grounding (RAG). There should also be meaningful and technically effective mechanisms to opt out of the use of their content.
A program corrected obvious text-recognition errors in this quote for display.
Ensure remuneration does not depend on publishers' negotiating ability.
Remuneration for content should not depend on a publisher's ability to negotiate, or be calculated solely based on volume and scale of content reach.
Require prominent attribution and links to source articles in licences for output-stage use.
This includes requiring prominent attribution and link to source articles on licenses covering output-stage use.
Maintain the decision not to introduce a text and data mining exemption and avoid loopholes under alternative wording.
The Government must also maintain the decision not to introduce a text and data mining exemption, and ensure no exceptions under alternative wording allow for a loophole which would give AI developers free access to Australian publishers’ content.
A program corrected obvious text-recognition errors in this quote for display.
Give consideration to news produced in audiovisual formats.
Consideration should also be given to news produced in audiovisual formats.
Read the original
In some quotes a program corrected obvious text-recognition errors for display.