AI Inquiry Atlas

An unofficial, independent guide to the public records of the Joint Select Committee on Artificial Intelligence. Not connected with the Parliament of Australia or the committee. Its makers also made a submission (no. 481; see About). Summaries and scores are written by AI and can be wrong: check the original before relying on them.

Submission 378

Name withheld

Submitter group: Individuals.

Position labels are not shown for individual submitters.

An AI model identified the submitter group and chose the extracts. Counts do not measure importance or merit.

Open in the interactive guide

Extracts from the submission, in the submitter’s own words.

Extracts

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.

  1. Data centres are often being treated like standard warehouses or infrastructure developments in industrial areas yet their impact is felt well beyond the fence line.

  2. Generative AI will reshape our workplaces; industry wide regulation, employee consultation and considered workplace protections are required to protect Australian workers from unintended impact from AI

  3. Dependency on generative AI as a coping mechanism or replacement for mental health care is a concerning new risk that has recently come to the attention of healthcare professionals69. Using generative AI conversational models is inherently addictive and is becoming an addiction seen in vulnerable adults, teenagers, and children

  4. It is understood that the government has implemented mandatory training in this space for all agencies, however this is not being enforced. The ‘AI In Government: Fundamentals Training’ is not consistently applied as mandatory training to all public servants using AI tools.

  5. generative AI models are rapidly reducing revenue streams for a number of creative and arts professions as well as education and wellbeing professions engaged in self-employed businesses.

  6. AI models have been and continue to be trained on large amounts of data through text and data mining (TDM) and ethically questionable labour practices51. This data is often copyrighted and this has created a significant legal challenge for the owners of intellectual property

Extracts from recommendations

Extracts chosen by an AI model, in the submitter’s own words. Listed in the same order as on its page in the interactive guide.

  1. Recommendation 1: Establish mandatory national standards and regulations for data centre renewable energy, diesel emissions limits and require the usage of treated wastewater for cooling.

  2. Recommendation 2: Establish user limitations and restricted access to generative AI for children and youth and require all major social media platforms, search engines, and applications operating within Australia, with integrated AI models and agents, to include mechanisms for users to opt-out of AI integration.

  3. Recommendation 3: Introduce mandatory community consultation and impact assessment of environmental and community wellbeing impacts for each new data centre application.

  4. Recommendation 4: Pause the adoption and integration of GovAI into government systems until there is a comprehensive workforce consultation and impact assessment.

  5. Recommendation 5: Establish a revenue levy on hyper scale data infrastructure and a destination tax on AI product companies, model and application providers, and AI deployment platforms.

  6. Recommendation 6: Establish a payroll-equivalent levy on private entities and government departments integrating AI into workflows.

  7. Recommendation 7: Establish compliance and evaluation of measures under the Policy for the responsible use of AI in government including comprehensive training requirements.

  8. Recommendation 8: Require all generative AI informed documents, video, imagery, products, and advertisements, to include a disclaimer statement for transparency and verification purposes.

  9. Recommendation 9: Ensure that all AI technology adoption within the education system and the workplace is an opt-in mechanism with no mandatory requirement for Australians to self-adopt or actively use AI models, systems or processes.

  10. Recommendation 10: Strengthen the current requirements and compliance around automated decision making within Commonwealth government, all decision making to be made by humans and have no AI influence.

  11. This feedback seeks to highlight the need for appropriate and genuine workplace consultation, training, meaningful guidance and comprehensive workplace policies to support the ethical adoption and implementation of generative AI across the country.

  12. There is no content on the environmental, health, social, and workplace risks and impact of generative artificial intelligence.

  13. It also refrained from mentioning that the training data for generative artificial intelligence was obtained through data scraping32, an action which lacks legal clarity and is currently the object of class action lawsuits and industry led action as well as public protest. Neither does it mention the exploitative and unethical labour practices that have been used to re ne data sets and train generative AI models33. These factors would inform how a public servant approaches adopting this tool in an ethical manner and should be included in the content of the module.

  14. The provision of adequate training and resources is a requirement under the Policy for the Responsible Use of AI in Government and should be a priority for government departments adopting and implementing generative AI models and tools in the workplace.

  15. Transitioning to renewable energy sources could be one way to reduce demand on the public energy grids36. Additionally, rostering limitations to alignment of high processing activity with lower grid demand periods could reduce strain on the power grid

  16. Regulations around emissions from data centres should be enforced as well as the requirement to adopt alternative options for backup energy sources such as renewable energy battery sources.

  17. Regulations around water consumption and usage should include a requirement to use treated effluent for cooling systems as well as the requirement to use air cooling methods or closed-loop cooling when possible. Limits on using town water supply, rainwater harvesting, and drawing from freshwater standing sources should also be applied.

  18. Current noise pollution regulations do not capture the low frequencies and consistency of data centre noise pollution and require review and updating to manage the noise emissions from data centres.

  19. Industry wide regulation, employee consultation and considered workplace protections are required to protect Australian workers from unintended redundancy from AI

  20. it is critical that users be aware and informed of the risk and are supported with proper reporting pathways, accountability, and liability, and be protected under relevant law.

  21. Current criminal legislation needs to be reviewed to address the particular instances of ambiguity surrounding accountability and responsibility with AI facilitated crime, psychological influence, and automated criminal activity through AI.

  22. Additionally, the current legislation around false advertising, fraud, vandalism and extortion, needs to be reviewed along with legal pursuit of enforcement against the definitions of accessory to a crime, affray, bullying, harassment, false or misleading information, false statements, falsification of documents, forgery, production of child pornography, threat to distribute intimate image, wilfully destroying records, unauthorised access to or modification of restricted data, unauthorised impairment of electronic communication, unauthorised impairment of data, which are all potential AI criminal activity that could be unprompted by a user or the result of altered code with no currently no accountable or liable individual.

  23. These laws require significant reform to support Australian creatives and intellectuals in protecting their work.

  24. Its important to consider how Australians and critical supply chains can be protected against the new threat of AI

  25. I implore you to consider the need for access and usage limitations on leading LLMs and AI models such as Gemini, CoPilot, OpenAI, ChatGPT, and Claude.

Read the original

Open in the interactive guide