AI systems

…and the challenges of governing AI internationally. Drawing on Lindroos-Hovinheimo’s work on AI regulation, the discussion connects national and European legal approaches with emerging questions about the governance and security of internet and AI systems.

The Decision Sciences and Analytics group recruits a doctoral researcher to study how AI systems should be designed around human decisions, particularly when consequences are high stakes or long term. Potential settings include clinical and legal decisions and education. The work is mainly theoretical, with experiments and external collaborations where needed, alongside teaching and student supervision. The initial appointment is 18 months, with extension toward a four- to five-year doctorate subject to progress. The start date is negotiable. Apply online with a one-page motivation letter, CV,

Build experimental biology programmes that help AI systems perform useful scientific research. Design molecular biology and biochemistry experiments, develop functional high-throughput screens and collaborate with computational researchers on AI-agent planning and evaluation. The role connects laboratory evidence with model capabilities. The advertised annual salary is USD 300,000–320,000. Anthropic’s hybrid policy requires at least 25% office attendance, with some roles requiring more.

…preparation and the sources of bias in AI systems. An external segment with Ian Scott and Ilaria Zanardi explores elephant communication, including low-frequency sound, long-distance signalling and the evidence for individually distinctive calls.

Supports multilingual and culturally contextual evaluation of AI systems, including benchmarks and datasets for underrepresented languages and contexts, with attention to the Global South. Proposals may develop responsible evaluation infrastructure, human judgment in evaluation, and methods for assessing AI deployment in emerging markets. Use the ARA proposal template; four pages excluding appendices are encouraged. Decisions are expected in February 2027. Applications close on 4 November 2026 at 11:59pm Pacific Time.

Jonathan Webb and ABC AI reporter Cam Wilson discuss how AI systems are changing mathematical research and the controversy around claims concerning the Navier–Stokes equations and the Millennium Prize. Mathematician Tristan Buckmaster of New York University contributes to the discussion.