Written by Sophie Babic

A $1B bet on the magic of AI 

Disney has announced a US$1 billion investment in OpenAI alongside a three-year licensing agreement.  

Positioned as an opportunity to empower fans, the deal enables users to generate short, prompt-based videos featuring more than 200 characters from Disney, Marvel, Pixar and Star Wars. A curated selection of this content will also be available on Disney+. 

The arrangement excludes actor likenesses and voices, avoiding additional rights and compensation issues, and includes guardrails restricting content involving topics such as sex, drugs and alcohol.  

While the partnership is only exclusive for one year, it signals a shift by major entertainment companies toward commercial collaboration with AI developers. 

Blueprint for Australia’s AI future 

On 2 December, the Australian Government released its National AI Plan, setting out a roadmap to build an inclusive AI economy, one that protects workers, addresses service gaps, and supports the growth of local AI capability. 

Australia will not introduce mandatory guardrails for high-risk AI systems. Instead, the government maintains that the existing legal and regulatory framework is largely sufficient, with any targeted reforms to address specific AI-related harms or risks to be supported by a new A$30 million AI Safety Institute within the Department of Industry. 

The National AI Plan outlines three core pillars to position Australia to compete globally, including: 

• Capturing opportunities by building smart infrastructure, backing domestic AI capability, and attracting global investment. 

• Spreading the benefits through widespread AI adoption, supporting and training Australian workers, and improved public services. 

• Keeping Australians safe with legislative and regulatory frameworks that mitigate AI harms, while promoting widespread responsible practices and international engagement that upholds Australia’s values. 

AI enters triple zero 

Could bots be in the near future for emergency services? A new study reveals Australians want artificial intelligence to improve the delivery of emergency services, with many willing to share personal data on Triple Zero (000) calls. 

The survey, by Researchscape for Motorola Solutions, found 58 per cent supported the implementation of AI that could detect vital information in emergency calls and accelerate a response. This includes keywords such as “knife” and “collision”, as well as translating live calls in foreign languages and processing live video to identify potential threats. 

While respondents overwhelmingly supported advanced technology to improve emergency services, they did not have confidence it would be used effectively. 

Real-world science vs AI 

OpenAI has launched FrontierScience, a benchmark that evaluates AI across physics, chemistry, and biology. It tests both Olympiad-style and real-world research tasks. 

Experiments with GPT‑5 show it can accelerate workflows, from literature searches to proofs and cross-disciplinary analysis. Tasks that took weeks can now take hours. 

In initial results, GPT‑5.2 scored 77 per cent on Olympiad-style questions and 25 per cent on research tasks. It outperformed other models and shows potential for structured reasoning in research. 

While FrontierScience helps track AI performance, highlighting strengths and areas for improvement, human oversight remains essential.