As part of the international expert advisory group representing Australia, Iโm pleased to see the second Key Update of the AI Safety Science Report, led by Yoshua Bengio, released (link in the comment). The official update captures the broad progress: more robust adversarial training, better AI-generated content tracking, wider adoption of frontier safety frameworks, and…
Most creators say they want protection from AI, yet a recent UK survey found over 90% would not opt out of AI access to their work. Why? Todayโs visibility depends on online discovery and search, and those systems are powered by AI. If you opt out, your work may be safe from AI training, but…
I often share missed or surprising observations in my talks. Here are five of them, summarised as food for thought. ๐ ๐๐๐ต ๐ญ: ๐ข๐๐ฒ๐ฟ๐๐ฒ๐ฎ๐-๐๐ฟ๐ฎ๐ถ๐ป๐ฒ๐ฑ ๐๐ ๐๐ผ๐ปโ๐ ๐ฎ๐น๐ถ๐ด๐ป ๐๐ถ๐๐ต ๐๐๐๐๐ฟ๐ฎ๐น๐ถ๐ฎ๐ป ๐๐ฎ๐น๐๐ฒ๐.Itโs intuitive to think globally trained models canโt reflect our norms. Yet when frontier models answer the same cultural questions posed to national cohorts, they align most…
Every confidential message youโve sent may already be stored โ waiting to be unlocked. This is the essence of the โ๐ต๐ฎ๐ฟ๐๐ฒ๐๐-๐ป๐ผ๐, ๐ฑ๐ฒ๐ฐ๐ฟ๐๐ฝ๐-๐น๐ฎ๐๐ฒ๐ฟโ problem: data intercepted today could be decrypted once quantum computers mature. This week CSIRO’s Data61 released our new report, ๐๐ช๐๐ฃ๐ฉ๐ช๐ข ๐๐๐๐ ๐๐ง๐๐ฃ๐จ๐๐ฉ๐๐ค๐ฃ: ๐๐๐๐ก๐๐ฉ๐ฎ, ๐๐ช๐ง๐๐ก๐๐จ ๐๐ฃ๐ ๐๐๐ฉ๐๐ฌ๐๐ฎ๐จ, mapping the scale of the challenge and…
What counts as a ๐ฑ๐ฒ๐ฐ๐ถ๐๐ถ๐ผ๐ป, really? From the strict sense in ๐ฎ๐ฑ๐บ๐ถ๐ป๐ถ๐๐๐ฟ๐ฎ๐๐ถ๐๐ฒ ๐น๐ฎ๐โwhere a decision formally alters someoneโs rights or entitlementsโto the everyday sense where almost everything, from selecting a movie to choosing a candidate, feels like a decision. Yet across these very different contexts, the underlying anatomy looks surprisingly similar. In CSIRO’s Data61 new…
Every few weeks brings a new AI framework โ but rarely one that genuinely helps organisations simplify. The National AI Centre‘s new guidance on responsible AI adoption, launched this week, is a welcome exception. It distils earlier guidance and complex international work into six clear, balanced practices that bridge governance intent with operational reality. We…
Cognitive overload, opaque automation, and AI surveillance are reshaping the modern workplace. Iโm sharing some insights from my talk at the Safety Sphere, Australiaโs leading HSE practitioner network, on โ๐ช๐ต๐ฒ๐ป ๐๐ ๐๐ฒ๐ฐ๐ผ๐บ๐ฒ๐ ๐ฎ ๐๐ผ๐๐ผ๐ฟ๐ธ๐ฒ๐ฟ: ๐จ๐ป๐ฑ๐ฒ๐ฟ๐๐๐ฎ๐ป๐ฑ๐ถ๐ป๐ด ๐๐ฆ๐ ๐ฅ๐ถ๐๐ธ๐ ๐ฎ๐ป๐ฑ ๐ข๐ฝ๐ฝ๐ผ๐ฟ๐๐๐ป๐ถ๐๐ถ๐ฒ๐ ๐ผ๐ณ ๐๐ฒ๐ป๐ฒ๐ฟ๐ฎ๐น-๐ฃ๐๐ฟ๐ฝ๐ผ๐๐ฒ ๐๐.โ We examined how AI not only enhances productivity but also creates new health,…
It was a great pleasure to deliver ๐ฎ ๐ธ๐ฒ๐๐ป๐ผ๐๐ฒ ๐๐ฒ๐๐๐ฒ๐ฟ๐ฑ๐ฎ๐ ๐ฎ๐ ๐๐ต๐ฒ ๐ฎ๐ณ๐๐ต ๐๐๐ ๐๐ป๐๐ฒ๐ฟ๐ป๐ฎ๐๐ถ๐ผ๐ป๐ฎ๐น ๐๐ผ๐ป๐ณ๐ฒ๐ฟ๐ฒ๐ป๐ฐ๐ฒ ๐ผ๐ป ๐ ๐๐น๐๐ถ๐บ๐ผ๐ฑ๐ฎ๐น ๐๐ป๐๐ฒ๐ฟ๐ฎ๐ฐ๐๐ถ๐ผ๐ป, ๐๐ต๐ฒ๐ฟ๐ฒ ๐ ๐ฒ๐ ๐ฝ๐น๐ผ๐ฟ๐ฒ๐ฑ ๐๐ต๐ฒ ๐ฒ๐๐ผ๐น๐๐ถ๐ป๐ด ๐ป๐ฎ๐๐๐ฟ๐ฒ ๐ฎ๐ป๐ฑ ๐ณ๐๐๐๐ฟ๐ฒ ๐ผ๐ณ ๐ต๐๐บ๐ฎ๐ป ๐ผ๐๐ฒ๐ฟ๐๐ถ๐ด๐ต๐. Recent reports such as OpenAIโs GDPVal and METR show that AI systems can now autonomously perform complex doing/solving tasks and often surpass human experts on…
It has been a great honour to represent Australia on the Expert Advisory Panel for the International AI Safety Report, led by Joshua Bengio and backed by over 30 countries and international organisations including the EU, OECD and UN. This is the most important ongoing effort to monitor AI capability gains and safety risks of…
๐ฅ๐ถ๐๐ธ-๐ฏ๐ฎ๐๐ฒ๐ฑ ๐๐ ๐ฝ๐ผ๐น๐ถ๐ฐ๐/๐ฟ๐ฒ๐ด๐๐น๐ฎ๐๐ถ๐ผ๐ป ๐๐ผ๐๐ป๐ฑ๐ ๐ผ๐ฏ๐๐ถ๐ผ๐๐. ๐๐๐ ๐ถ๐ ๐ฟ๐ฒ๐๐๐ ๐ฎ ๐ณ๐ฟ๐ฎ๐ด๐ถ๐น๐ฒ ๐ฎ๐๐๐๐บ๐ฝ๐๐ถ๐ผ๐ป: ๐๐ต๐ฎ๐ ๐๐ฒ ๐ฐ๐ฎ๐ป ๐ฎ๐ฐ๐๐๐ฎ๐น๐น๐ ๐บ๐ฒ๐ฎ๐๐๐ฟ๐ฒ ๐ฟ๐ถ๐๐ธ.Yet risk assessment is often put forward as the first thing to answer, as if itโs the easy part. Some frameworks push โuse caseโbasedโ solutions with a pre-defined list of high-risk use cases, but then smuggle in the magic…
Research Director, CSIRO
Conjoint Professor, CSE UNSW
For other roles, see LinkedIn & Professional activities.
If you’d like to invite me to give a talk, please see here & email liming.zhu@csiro.au