Australia Gains Access to Anthropic's Mythos AI

Anthropic has granted 150 organizations across 15 countries, including Australia, access to its Claude Mythos Preview as part of an expansion of Project Glasswing. The model, which can identify software and network vulnerabilities, was previously limited to UK and US governments and about 50 US companies due to its cybersecurity capabilities. Australia's federal government declined to confirm details of its access, while the nation's cyber security agency welcomed the inclusion and independent MP Kate Chaney urged the government to accelerate its response to AI risks.

Australia Gains Access to Anthropic's Mythos AI ABC News reports that Australia and several private companies are among 150 organisations across 15 countries granted access to Anthropic's Claude Mythos Preview as part of an expansion of Project Glasswing. ABC reports the model has been withheld from broad public release because of its cybersecurity capabilities; the company previously limited access to UK and US governments and about 50 US companies, ABC reports. The federal government declined to confirm details of its access, while the nation's cyber security agency welcomed the inclusion, ABC reports. Independent MP Kate Chaney is urging the government to accelerate its response to AI risks and opportunities, ABC reports. ABC also cites a new report warning that an AI-driven surge in data centres in Australia threatens the country's energy transition. What happened ABC News reports that Anthropic has granted 150 organisations across 15 countries , including Australia, access to its Claude Mythos Preview as part of an expansion of Project Glasswing. ABC reports the model has not been released publicly because of its cybersecurity capabilities. According to ABC, the model is able to identify software and network flaws that would pose a significant cybersecurity risk in the wrong hands. ABC reports Anthropic had previously limited Mythos access to the UK and US governments and to about 50 US companies. ABC reports Australia's federal government declined to confirm specifics of its access, and that the nation's cyber security agency welcomed the inclusion. ABC reports independent MP Kate Chaney has called on the government to step up its response to AI, and ABC cites a new report warning that AI-driven demand for data centres in Australia jeopardises the country's energy transition. Technical details ABC reports that Claude Mythos Preview is capable of surfacing software and network vulnerabilities, a capability the article frames as the reason for restricted distribution. Editorial analysis - technical context: Powerful models that can automatically identify exploitable code and network weaknesses are an example of dual-use AI, presenting both defensive benefits and potential offensive misuse. For practitioners, this raises operational questions about access controls, red-team practices, and secure deployment pipelines when working with vulnerability-detection models. Context and significance Public reporting frames this expansion of access as part of broader tensions between rapid capability development and governance readiness. Observers following the sector increasingly note that restricted or staged model rollouts aim to balance research and security, while regulators and procuring agencies face new pressure to update procurement criteria, oversight, and incident-response playbooks. What to watch - •Whether Australian agencies or companies disclose usage policies, oversight arrangements, or contracted safeguards for restricted-model access, as these will shape precedents for future procurement. - •Any technical disclosures from Anthropic about access controls or deployment constraints for Claude Mythos Preview, which will affect how practitioners integrate such tools into defensive workflows. - •Follow-up reporting on the cited data-centre report and government energy planning, since rising compute demand can materially affect infrastructure and sustainability planning. Scoring Rationale The story matters because restricted access to a model that can find software and network flaws raises practical security and governance questions for practitioners, but it is not a broadly released model or a paradigm-shifting publication. The immediate impact is notable for security teams, procurement officers, and policy watchers. Practice interview problems based on real data 1,500+ SQL & Python problems across 15 industry datasets — the exact type of data you work with. Try 250 free problems /problems