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 115

Soroptimist International Australia Inc

Submitter group: Civil society.

An AI model identified the submitter group and wrote the summary, key points and recommendation summaries. Counts do not measure importance or merit.

Open in the interactive guide

Summary

Read the original submission for the full text.

Argues for a human-centred approach to AI governance built on safety, equality, accountability and sustainability by design. Sets out recommendations on gendered AI harms and digital safety, accountable automated decisions, women's workforce transition and creator protections, and the resource and data sovereignty impacts of AI infrastructure.

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.

  1. Argues that responsibility should rest primarily on those who design, develop, deploy and profit from AI systems rather than on individuals expected to protect themselves after harm.

    A human-centred approach should therefore place responsibility primarily on those who design, develop, deploy and profit from AI systems—not on individuals expected to protect themselves after harm has occurred

  2. Says women and girls can face risks from sexualised deepfakes, image-based abuse, impersonation, harassment and coercive control.

    Women and girls can face risks from sexualised deepfakes, image-based abuse, impersonation, harassment, coercive control and reputational harm.

  3. Recommends that consequential automated decisions include human oversight and rights to explanation, review and appeal.

    Requiring meaningful human oversight for consequential decisions, with a clear right to explanation, review and appeal.

  4. Recommends AI literacy and STEM investment for women and girls, including in regional, remote and disadvantaged communities.

    Targeted investment in AI literacy, digital capability and STEM pathways for women and girls, with specific attention to regional, remote and disadvantaged communities.

  5. Argues that the AI Safety Institute should support regulators with testing and evidence rather than act as the primary regulator.

    The Australian AI Safety Institute (AISI) should support this regulatory ecosystem through testing, technical analysis and evidence on emerging risks. It should not be treated as the primary regulator

Recommendations

Recommendations are shown beside the submission’s own words. Listed in the same order as on its page in the interactive guide.

  1. Adopt a human-centred, risk-based, enforceable AI governance framework combining safeguards with support for innovation, workforce transition, Australian capability and responsible investment.

    S| Australia therefore calls for an AI governance framework that is human-centred, risk-based, and enforceable. It should combine strong safeguards for people with measures that support innovation, workforce transition, Australian capability and responsible investment (Australian Government Department of Industry, Science and Resources, 2025; UNESCO, 2025).

    A program corrected obvious text-recognition errors in this quote for display.

  2. Build AI governance on safety before deployment; evaluation of high-risk systems for discrimination and gendered/intersectional impacts; information, review and appeal for consequential decisions; and assessment of energy, water, infrastructure and community impacts alongside benefits.

    We recommend that Australian AI governance be built around four principles: • Safety by design: foreseeable risks should be identified and mitigated before systems are deployed. • Equality by design: high-risk systems should be evaluated for discrimination and unequal impacts, including gendered and intersectional impacts. • Accountability by design: people affected by consequential automated decisions should have meaningful information, human review and avenues for appeal. • Sustainability by design: the energy, water, infrastructure and community impacts of AI should be assessed alongside its economic benefits.

  3. Establish statutory safety duties for developers and deployers of high-risk generative AI systems, supported by risk assessments, testing, monitoring and documented mitigation.

    Establishing clear statutory safety duties for developers and deployers of high-risk generative AI systems, supported by risk assessments, testing, monitoring and documented mitigation measures.

  4. Strengthen technical and platform safeguards, resistant to circumvention, against generating and distributing non-consensual intimate imagery and other clearly harmful synthetic content.

    Strengthening technical and platform safeguards to prevent the generation and distribution of non- consensual intimate imagery and other clearly harmful synthetic content, with safeguards designed to resist circumvention.

  5. Ensure rapid, accessible removal pathways for AI-generated image-based abuse, with strong compliance requirements and meaningful penalties for systemic failure.

    Ensuring rapid and accessible removal pathways for AI-generated image-based abuse, with strong compliance requirements and meaningful penalties for systemic failure.

  6. Ensure eSafety and other relevant regulators have the powers, resources and technical capability to respond to AI-enabled harms.

    Ensuring eSafety and other relevant regulators have the powers, resources and technical capability required to respond to rapidly evolving AI-enabled harms.

  7. Have the Australian AI Safety Institute support the regulatory ecosystem through testing, technical analysis and evidence on emerging risks, without being treated as the primary regulator.

    The Australian AI Safety Institute (AISI) should support this regulatory ecosystem through testing, technical analysis and evidence on emerging risks. It should not be treated as the primary regulator

  8. Require proportionate Gender and Diversity Impact Assessments and other risk assessments for high-risk AI and automated decision-making systems, covering discrimination, accessibility, privacy and foreseeable social impacts.

    Requiring proportionate Gender and Diversity Impact Assessments and other risk assessments for high- risk AI and automated decision-making systems, covering discrimination, accessibility, privacy and foreseeable social impacts.

  9. Require meaningful human oversight of consequential decisions, with a clear right to explanation, review and appeal.

    Requiring meaningful human oversight for consequential decisions, with a clear right to explanation, review and appeal.

  10. Require organisations deploying high-risk AI to keep records of system purpose, data sources, testing, performance, limitations, incidents and mitigation.

    Requiring organisations deploying high-risk AI to maintain appropriate records of system purpose, data sources, testing, performance, known limitations, incidents and mitigation measures.

  11. Strengthen privacy protections for AI-related profiling, tracking and data brokerage, including considering the continued operation of the Privacy Act small-business exemption (turnover $3 million or less, subject to existing exceptions).

    Strengthening privacy protections for AI-related profiling, tracking and data brokerage, including consideration of the continued operation of the Privacy Act’s small-business exemption. The exemption applies to businesses with annual turnover of $3 million or less, subject to existing exceptions.

  12. Ensure public-sector AI use is transparent, contestable and subject to clear lines of accountability.

    Ensuring public-sector AI use is transparent, contestable and subject to clear lines of accountability.

  13. Invest in AI literacy, digital capability and STEM pathways for women and girls, with attention to regional, remote and disadvantaged communities.

    Targeted investment in AI literacy, digital capability and STEM pathways for women and girls, with specific attention to regional, remote and disadvantaged communities.

  14. Fund reskilling and career-transition pathways for workers in occupations significantly changed by AI, including accessible training and recognition of existing skills.

    Funded reskilling and career-transition pathways for workers in occupations experiencing significant AI- related change, including accessible training and recognition of existing skills.

  15. Collaborate with employers, unions, education providers and industry to support job quality, worker voice and fair transition.

    Collaborate with employers, unions, education providers and industry to support job quality, worker voice and fair transition as AI is introduced.

  16. Strengthen protections for creative workers and creators whose works, voices, images, likenesses or other IP are used in AI development without appropriate authorisation, attribution or compensation.

    Strengthen protections for creative workers and creators where their works, voices, images, likenesses or other intellectual property are used in AI development without appropriate authorisation, attribution or compensation.

  17. Require major AI and data-centre developments to undertake transparent assessments of energy, water, environmental and community impacts, including cumulative impacts.

    Requiring major AI and data-centre developments to undertake transparent assessments of energy, water, environmental and community impacts, including cumulative impacts.

  18. Promote renewable energy procurement, energy efficiency, demand flexibility and lower-water cooling where technically and economically feasible.

    Promoting renewable energy procurement, energy efficiency, demand flexibility and lower-water cooling approaches were technically and economically feasible.

  19. Consider mechanisms so communities hosting high-density digital infrastructure share in benefits and do not absorb disproportionate environmental or infrastructure costs.

    Considering mechanisms to ensure that communities hosting high-density digital infrastructure share in benefits and are not left to absorb disproportionate environmental or infrastructure costs.

  20. Protect Australian data sovereignty and recognise First Nations Data Sovereignty in the development, storage, use and governance of AI datasets and systems.

    Protecting Australian data sovereignty and recognising First Nations Data Sovereignty in the development, storage, use and governance of AI datasets and systems.

  21. Integrate AI infrastructure planning with broader energy, water, land-use and critical-infrastructure planning.

    Ensuring AI infrastructure planning is integrated with broader energy, water, land-use and critical- infrastructure planning.

Read the original

In some quotes a program corrected obvious text-recognition errors for display.

Open in the interactive guide