UNSW AI Institute

Meet our AI Program Managers Kate Gwynne, PhD & Shetal B. Kate works at the intersection of research, innovation, and human-centred approaches to cutting-edge technologies. She holds a PhD in emerging technology from UNSW’s School of Arts and Media and has led innovation initiatives for global organisations in marketing and advertising, as well as advising startups and local business leaders. As Practical AI Program Manager at UNSW Founders, her goal is to empower UNSW startups and the broader community to embrace AI tools and technologies – helping them supercharge their capabilities, achieve their goals, and do it all with less friction and more fun. Shetal is the Practical AI Program Manager at UNSW Founders, driven by the belief that AI should serve people, not just impress them. With a background in physics, computer science, and science communication, she develops AI systems and delivers workshops that make AI ethical, accessible, and impactful. Her mission: to make AI make sense – and make a difference.

New from me: When Australian AI expert Dr Kobi Leins (PhD, GAICD) declined a medical specialist’s request to use AI transcription software during her child’s upcoming appointment, the practice told her the practitioner required the technology be used and Leins was "welcome to seek an alternative" provider. 👋 The specialist’s “high workload and limited time” meant they were now “unable to perform timely assessments without the use of AI transcribing tools”, the practice said in an email seen by Information Age. “They gave me the choice of proceeding with the AI that they were insisting on, or to find another practitioner,” Leins said. In Australia, healthcare providers can decide to not consult with a patient for any reason — except in emergency situations — provided they facilitate the patient’s ongoing care in some way. The system used by the specialist was one whose privacy and security capabilities Leins had previously reviewed as part of her work in AI governance — and one she said she would not want her child’s data “anywhere near”. And unlike most medical devices, AI scribes remain largely unregulated, leaving it up to individual healthcare practices to decide which tools they want to use, and how. The University of Queensland associate professor Dr Saeed Akhlaghpour said regulating AI scribes in healthcare with “targeted, risk-based national oversight” would help prevent potential issues around privacy, security, and liability. Full story: https://lnkd.in/gCJKR5Ne

Kudos to the Productivity Commission for spotlighting some of the most pressing issues in AI and digital governance in its comprehensive interim report on harnessing data and digital technology to boost Australia’s productivity. But as someone deeply focused on safe, sovereign & strategic AI - and Australia’s broader strategic digital capability - I believe the proposals need refinement to truly serve our national interest: - Re: AI Regulation: The PC recommends pausing mandatory guardrails and relying on existing laws. However, frontier AI models introduce novel risks that legacy frameworks can’t fully address. Let’s establish an Australian AI Safety Institute and implement tiered guardrails for high-risk models, aligned with global best practice. - Re: Copyright & AI Training: We need to be very cautious of downplaying the risk of undermining Australia’s copyright system. One option is to introduce a TDM exception for research and non-commercial use to empower universities, medtech, and startups to build ethically and responsibly on Australian data. - Re: Data Access Reform: The proposed tiered pathways are promising but risk fragmentation without stronger coordination. The creation of a National Data Access Framework, with baseline standards and public-private intermediaries, would drive real impact. - Re: Privacy Reform: The dual-track compliance model is a smart idea, but needs clearer enforcement and metrics. I'd suggest developing pilots in low-risk sectors, and establishing a Board to guide implementation and build trust. - Re: Digital Financial Reporting: The mandate is welcome, but a bit too narrow. I'd expand it to include large proprietary companies and sustainability reporting, and build-in capability to enable AI-powered analysis - such as by adopting iXBRL. #AI #Productivity #Australia #DigitalTransformation #Innovation #AIRegulation

Save the date - 28 August. Not long ago, I wrote a post about the National AI Centre's Australian AI Ecosystem Report: https://lnkd.in/g3aJ5QhT The headline summary was that Australia actually produces top-tier AI research but struggles to commercialise them due to structural barriers like limited VC, weak manufacturing, high business costs, etc. So Australia has an 'evolutionary' (rather than 'revolutionary') and 'specialised regional' approach to AI, by focusing on applications in existing strengths of resources, energy, biotech, finance, govt/defence - each of which is driven from a certain city/region hub. The report inspired me to dig deeper in Australia's AI ecosystem and policy mix - so much so that I thought "hey, why not ask the National AI Centre directly?" 🇦🇺 And so here we are. Proud to announce that my association (Data Science and Ai Association of Australia | DSAi) will be partnering with National AI Centre x UNSW Founders x UNSW AI Institute to deliver a Q&A fireside chat with eminent guests: ⭐ Lee Hickin - Executive Director of National AI Centre ⭐ Sebastian Sequoiah-Grayson - Science Outreach Officer of UNSW AI Institute and Senior Lecturer in Epistemics at UNSW Computer Science and Engineering I'll be moderating one of the Q&A sessions, and hope to dive into the global state of AI, the ambition for Australia and some of the practical considerations we need to adapt in order to build success as a country. Because we'll only have 2 speakers, we'll have the time to go real deep (beyond the generic stuff). If you're based in Sydney, just come along! Hope to meet (and catch up) with as many of you as possible :) It's a free event btw. Pizzas + networking opportunities. Our audiences are typically startups, VC investors, developers, product managers, and professionals. Registration link in the comments below. #australia #ai #tech #event #chat

Who should pay for the data that trains AI models? The recent Productivity Commission suggestion to relax copyright laws for AI training has reignited debate over how Australia should balance AI innovation with protecting its creative and research sectors. In an article from The Australian Financial Review, Kingston AI Group member and UNSW AI Institute Chief Scientist Professor Toby Walsh FAA FTSE FRSN is clear: breaching copyright isn’t necessary to build effective AI. “There’s absolutely no need for [AI companies] to do it other than they’re in a competitive race with each other,” he said. Professor Walsh points out that major AI companies have already signed licensing deals with news organisations and publishers. “The fact that they’re doing deals [with publishers] demonstrates that they should have paid those licensing deals in the first place. I don’t buy the argument that this is fair use.” In a submission to Federal Treasurer Jim Chalmers, UNSW’s AI Institute warned that exempting AI from copyright and IP laws: “…would amount to a massive, uncompensated transfer of value from Australia’s creative and research sectors to, predominantly, foreign technology companies, undermining our cultural sovereignty and reinforcing their monopolies,” the submission said. As Australia considers its place in the AI race, voices like Professor Walsh’s remind us that control over our training data is the foundation of AI sovereignty. To read the article from the AFR, please visit: https://lnkd.in/gcxTrezw Photo caption: UNSW AI Institute Chief Scientist Professor Toby Walsh. Photo credit: UNSW

𝗙𝗿𝗲𝘀𝗵 𝗶𝗱𝗲𝗮𝘀. 𝗨𝗻𝗹𝗶𝗸𝗲𝗹𝘆 𝗮𝗹𝗹𝗶𝗲𝘀. 𝗣𝗿𝗮𝗴𝗺𝗮𝘁𝗶𝗰 𝘀𝗼𝗹𝘂𝘁𝗶𝗼𝗻𝘀. Last night in Sydney, with 𝗦𝗲𝗻𝗮𝘁𝗼𝗿 𝘁𝗵𝗲 𝗛𝗼𝗻 Tim Ayres, 𝗠𝗶𝗻𝗶𝘀𝘁𝗲𝗿 𝗳𝗼𝗿 𝗜𝗻𝗱𝘂𝘀𝘁𝗿𝘆 𝗮𝗻𝗱 𝗜𝗻𝗻𝗼𝘃𝗮𝘁𝗶𝗼𝗻, 𝗠𝗶𝗻𝗶𝘀𝘁𝗲𝗿 𝗳𝗼𝗿 𝗦𝗰𝗶𝗲𝗻𝗰𝗲, we launched TPDi’s 𝗹𝗮𝘁𝗲𝘀𝘁 𝗿𝗲𝘀𝗲𝗮𝗿𝗰𝗵 𝗿𝗲𝗽𝗼𝗿𝘁, 𝘛𝘦𝘵𝘳𝘪𝘴 𝘧𝘰𝘳 𝘈𝘶𝘴𝘵𝘳𝘢𝘭𝘪𝘢’𝘴 𝘍𝘶𝘵𝘶𝘳𝘦: 𝘈𝘭𝘪𝘨𝘯𝘪𝘯𝘨 𝘖𝘶𝘳 𝘕𝘢𝘵𝘪𝘰𝘯𝘢𝘭 𝘈𝘐 𝘗𝘳𝘪𝘰𝘳𝘪𝘵𝘪𝘦𝘴 and formally announced the 𝗧𝗲𝗰𝗵 𝗣𝗼𝗹𝗶𝗰𝘆 𝗗𝗲𝘀𝗶𝗴𝗻 𝗙𝘂𝗻𝗱 a first-of-its-kind multistakeholder initiative backing independent Australian tech policy research. The diversity of Fund contributors – government, Aussie companies, multinational companies, professional associations and philanthropy - who don't always agree - reflects the strong appetite for fresh ideas and evidence-based thinking on how we want technology to shape Australia’s future. TPDi's Founding Fund Sponsors are: 𝗣𝗹𝗮𝘁𝗶𝗻𝘂𝗺 · Amazon · Commonwealth Bank · Australian Government Department of Finance · Minderoo Foundation · 𝗚𝗼𝗹𝗱 · ACS (Australian Computer Society) · Department of Industry, Science and Resources · Qoria · Microsoft · 𝗦𝗶𝗹𝘃𝗲𝗿 · Adobe · Apple · Atlassian · Macquarie Technology Group · Salesforce Learn more about the Fund 🔗 link in comments TPDi's new report shows how the different elements of national AI policy - people and the planet, productivity, trusted information and institutions, regulatory courage, global governance and national capability - can either fit together or clash like Tetris pieces. Our report provides a clear framework to help Australia align these moving pieces, clicking them into place by establishing a 𝗠𝗶𝗻𝗶𝘀𝘁𝗲𝗿𝗶𝗮𝗹 𝗔𝗜 𝗧𝗮𝘀𝗸𝗳𝗼𝗿𝗰𝗲 with a mandate to develop – and immediately execute - a 𝗡𝗮𝘁𝗶𝗼𝗻𝗮𝗹 𝗔𝗜 𝗦𝘁𝗿𝗮𝘁𝗲𝗴𝘆 𝗮𝗻𝗱 𝗔𝗰𝘁𝗶𝗼𝗻 𝗣𝗹𝗮𝗻. Read the report: Tetris for Australia’s Future 🔗 Link in comments We also had a great debate: more on this soon. But for now thanks to our wonderful debaters: Sue Keay · Gisele Kapterian · Tom Rogers AO · Tobias Feakin · Lizzie O'Shea FAAL · Sally-Ann Williams FTSE · Minh Hoang · Ian Oppermann And thank you to Sassoon Grigorian and the incredible team at Salesforce for hosting us at your beautiful venue. Johanna Weaver · Zoe Jay Hawkins · Meredith Hodgman · Dorina Wittmann

How do I stay across Australian AI policy activity? Easy. Everything goes into one of my AI notebooks. The Australian Productivity Commission yesterday published the interim report for its inquiry “Harnessing data and digital technology.” It's an interesting 124-page read and contains seven draft recommendations; and with it, comes a new round of consultation until 15 September. The inquiry has attracted some enthusiastic lobbying. But working out who the players are, what they want and why, isn’t always straightforward. But all the public consultation docs (≈115 lobbying submissions) and the interim report are now in a Google NotebookLM collection. [Update: 7 Aug. I've removed the public link, as Google NotebookLM doesn't allow different access levels for public / shared accounts. Send me a message with your Gmail address and I can share it with you] Some great questions to ask: ▪️Who supports or opposes a copyright exception for AI training (text and data mining). Why? ▪️Which stakeholders are calling for changes to privacy law; especially around targeted advertising, worker data, or facial recognition? ▪️What are the key tensions between Big Tech and Australia’s creative industries? ▪️What are the main targets for criticism? ▪️Who will be most opposed to the interim report's draft recommendations? Stuck for ideas? Ask it for a summary and then "What's a good question?"

AI is set to become the single biggest driver of global energy demand. The International Energy Agency (IEA) forecast electricity use from data centres is set to more than double by 2030. That’s more than the total consumption of Japan. That was the backdrop for Nuclear Futures: Shaping Dialogue in a Changing World, hosted this week by UNSW and the OECD Nuclear Energy Agency. I joined global experts and emerging leaders to tackle one of the defining energy challenges of our time: how to meet exponential energy demand with clean, firm, cheap energy. At the event, UNSW Engineering launched Australia’s first standalone Bachelor of Nuclear Engineering. A four-year program starting in 2026. It will train the engineers who will power AI infrastructure, defence, nuclear medicine and a low-emissions economy. I discussed the key drivers behind the world’s leading companies speed and scale on nuclear: - Microsoft is restarting Three Mile Island to secure 835 MW of clean power for its AI data centres - Google has signed a multi-gigawatt SMR deal targeting over 10 GW of zero-emissions capacity by 2035 - Amazon has invested over 500 million dollars into nuclear for its AWS infrastructure in the United States This is the new energy economy. And Australia has an amazing opportunity to expand its capability as a global leader in many aspects of nuclear yet is still banned from taking advantage of the energy domestically while the world races forward. Australia has: - One of the world’s largest uranium reserves - A national radioactive waste facility at Sandy Ridge - A skilled workforce scaling to support AUKUS nuclear submarines - A nuclear engineering degree to grow next-generation capability - ANSTO and the Opal reactor, a global leader in nuclear medicine ANSTO produces over 80% of Australia’s nuclear medicine and exports medical isotopes to more than 45 countries. The future of global energy is not one technology. It’s all of them. Solar. Wind. Battery. Nuclear. That’s the portfolio that will drive the energy transition, reducing emissions while keeping power reliable and affordable. Lots of great questions during our panel sessions: Why do we need a national decision on nuclear? Because it’s the only zero-emissions technology banned by law. We don’t need subsidies. We need permission to compete. Can nuclear be depoliticised? Yes. It’s already bipartisan in the US and across Europe. Australia can lead that same shift, if we choose to. How do we move past pro versus anti? Solve the energy trilema: green, firm and low cost electrons. Talk about the science not the political history. Stop telling people it’s too expensive and takes too long - get out of the way and let the free market operate. Lower emissions. National resilience. Energy security. The future of energy is not either-or. It’s everything. Every industrial revolution has been powered by energy. This one will be no different. #ai #digitalinfrastructure https://lnkd.in/gC_Vn7UJ

Our network member, Robotics and Mechatronics Society UNSW has officially launched its 8-week Sumobots Robotics Competition — unofficially the largest student-run robotics competition in the southern hemisphere! This year, over 300 students from multiple universities and disciplines have signed up to design, build, and battle their custom-built sumobots. The competition will culminate in a live knockout finale on August 1, followed by an Industry Night Trade Show — open to all companies and free to attend. This initiative is completely extra-curricular, with no course credits offered — showcasing some of the most motivated and proactive Mechatronic Engineering students at UNSW. Interested in learning more or attending the finale? Contact: info@ramsocunsw.org Follow us for regular updates about the robotics industry and community in Australia – thanks for your support! #robotics #innovation #automation #supportroboticsinaustralia