TL;DR
Australia's autonomous Ghost Bat jet fired and destroyed an aerial target with an air-to-air missile in a test over South Australia, marking a milestone in AI-assisted lethal autonomous systems.
What happened
- Ghost Bat, an Australian-made unmanned jet roughly two-thirds the size of a fighter, shot down an aerial target with an air-to-air missile in a test over remote South Australia in late 2024.
- The aircraft is not AI-piloted in the generative sense: Air Marshal Stephen Chappell confirmed it runs on deterministic programming, choosing from thousands of pre-designed options rather than improvising.
- A human operator aboard an E7 Wedgetail aircraft gave the Ghost Bat the order to engage, and then issued a final authorisation before the missile was released.
- Defence Minister Richard Marles separately traveled to the United States to meet with OpenAI, Anthropic, Microsoft, and Google to assess where frontier AI models are heading for defence use.
- The ADF has also adopted Palantir's Maven Smart system for intelligence analysis, drawing controversy over founder political ties and data sovereignty concerns.
Why it matters
- The test confirms Australia as an active participant in the autonomous weapons race, alongside Ukraine (AI-piloted drones targeting Russian supply routes) and Israel (the Lavender targeting system flagging 37,000 suspected operatives).
- The human-in-the-loop design is deliberate and legally significant: Chappell stressed full predictability of outcomes as essential for legal and ethical compliance, setting a template other democracies may follow or contest.
- The UN and ICRC issued a joint call last month to ban autonomous weapons that are unpredictable or that autonomously target humans, putting Ghost Bat-style programs directly in the crosshairs of international humanitarian law debates.
- Experts at ANU argue the bigger AI shift is upstream: intelligence fusion and decision-support tools like Maven Smart, not the weapons themselves, are where AI is reshaping command most rapidly.
- Israel's exponentially expanded AI use in Gaza, per IISS analysis, shows that faster targeting increases error risk when human assessment cannot keep pace, a cautionary signal for any force scaling these systems.
What to watch next
- Whether Australia formalises its human-authorisation doctrine into binding policy or treaty commitments, especially as UN pressure for an autonomous weapons ban intensifies.
- The Ghost Shark autonomous submarine program, the underwater counterpart to Ghost Bat, as the next visible test of ADF autonomous lethality at sea.
- How OpenAI, Anthropic, and Google respond to Marles's outreach: any announced defence partnerships would signal a major shift in how frontier AI labs engage with lethal applications.
Originally published on Present of AI, a daily source-linked AI news timeline. Read the full timeline or browse the open dataset.