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 303

Woolworths Group

Submitter group: Companies.

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.

Describes Woolworths' deployment of AI in customer experience, supply chains and workforce programs, and argues for whole-of-economy settings that support adoption and upskilling. Opposes AI-specific security legislation, supports outcomes-based, proportionate regulation, and proposes allocating responsibility along the AI supply chain by context rather than a fixed developer and deployer split.

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. Says its customer-facing, low-risk AI tools can lift shopping convenience and deliver more value without creating systemic risk.

    These practical applications illustrate how customer-facing, low-risk, AI-enabled tools can tangibly lift shopping convenience and deliver more value to customers, without creating systemic risk.

  2. Says its experience is that retail automation shifts roles toward higher-value work and creates different employment opportunities.

    Our experience has demonstrated that retail automation creates different employment opportunities by shifting roles toward higher-value, more rewarding work.

  3. Says AI-specific security legislation is not required and that existing frameworks including SOCI are fit for purpose when applied and enforced consistently.

    We do not view that AI-specific security legislation is required to manage these risks. The existing frameworks, including SOCI, are fit for purpose when applied and enforced consistently.

  4. Argues that responsibility should be allocated by context and by who is best placed to manage the risk, rather than a fixed developer and deployer split.

    Rather than codifying a fixed developer and deployer split, we think a more durable framework is one that allocates responsibility based on the context and nature of the decision being made and who is best placed to manage its associated risk.

  5. Using Agentic Olive as an example, says Google builds and controls the foundation model and Woolworths can only set application-specific guardrails.

    Olive is powered by Google Cloud’s Gemini Enterprise, meaning the foundational model and its underlying properties are built and controlled by Google and Woolworths can only set the application-specific guardrails (to be implemented by Google) to constrain the way Olive interacts with our customers.

  6. Says it is positioned as a national anchor for sovereign AI capability and that developing and retaining domestic talent and IP is critical.

    we are uniquely positioned as a national anchor for sovereign AI capability. Developing and retaining domestic talent and IP is critical to Australia’s long-term economic position.

Recommendations

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

  1. Achieve whole-of-economy policy settings that enable greater AI adoption by Australian businesses, invest in upskilling and reskilling Australians and helping them transition, and keep Australian businesses competitive against overseas peers.

    we think the Government’s focus should be on achieving whole-of-economy policy settings that: ● enable greater AI adoption by Australian businesses, which will create opportunities for new jobs locally; ● invest in upskilling and reskilling Australians and helping them transition as the economy shifts; and ● keep Australian businesses competitive against their overseas peers who are adopting technology at pace and are quickly building presence in Australian markets - this is not just about ‘AI policy’.

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

  2. Continue regulating security, software integrity and resilience for essential food and grocery supply chains under SOCI rather than AI-specific legislation, avoiding broad 'critical infrastructure' definitions in AI-specific regulation that would capture routine retail tools.

    Capturing routine retail operational tools under broad definitions such as "critical infrastructure" in AI-specific regulation would impose compliance burden without enhancing security outcomes and risks duplicating obligations that are already being met under purpose-built frameworks. Instead security, software integrity and resilience settings for essential food and grocery supply chains should continue to be regulated under SOCI, ensuring critical national assets are protected without creating duplicative regulation.

  3. Establish an AI regulatory framework with practical boundaries that regulate outcomes rather than underlying software code.

    Australia’s AI regulatory framework must establish practical boundaries that regulate outcomes rather than underlying software code.

  4. Place transparency and disclosure obligations with whoever is best placed to give meaningful, contextually relevant information to the customer, rather than assigning them rigidly across the supply chain.

    The same logic applies to transparency and disclosure obligations, which should sit with whomever is best placed to provide meaningful and contextually relevant information to the customer, rather than being assigned rigidly across the supply chain.

  5. Allocate responsibility for AI based on the context and nature of the decision and who is best placed to manage the risk, rather than codifying a fixed developer and deployer split.

    Rather than codifying a fixed developer and deployer split, we think a more durable framework is one that allocates responsibility based on the context and nature of the decision being made and who is best placed to manage its associated risk.

  6. Avoid additional AI-specific definitions or regulation where each AI supply-chain party meets its existing obligations under applicable law.

    Neither framework currently delineates responsibility across the AI supply chain, but where each party in that chain meets their existing obligations under applicable law, additional AI-specific definitions or regulation should not be necessary.

Read the original

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

Open in the interactive guide