Defence's advanced artificial intelligence projects have a new set of guidelines for ethical and legal considerations.
Subscribe now for unlimited access.
or signup to continue reading
Tools, which include a checklist and risk measurement system, are there to "operationalise ethical AI" in Defence projects that potentially have significant implications for human safety and other ethical issues.
But it goes far beyond the concept of killer robots that many may imagine.
Chief Defence Scientist Tanya Monro said AI technologies offer many benefits such as saving lives by removing humans from high-threat environments and improving Australian advantage by providing more in-depth and faster situational awareness.
"Upfront engagement on AI technologies, and consideration of ethical aspects needs to occur in parallel with technology development," Professor Monro said.
![Chief Defence Scientist Tanya Monro says ethics and technology must be considered in parallel. Picture: Defence Chief Defence Scientist Tanya Monro says ethics and technology must be considered in parallel. Picture: Defence](/images/transform/v1/crop/frm/gbZxCg3zJpb4r79EPiJSKy/18569c19-40ae-45b7-b787-390d5d7467b6.jpg/r0_24_1200_739_w1200_h678_fmax.jpg)
A workshop in 2019 on how Defence should incorporate ethical considerations into its projects turned into a collaboration between government, industry, academics, ethicists and media, including this journalist.
A paper published on Tuesday by the Defence Science and Technology Group based on that collaboration proposed a checklist and risk matrix that are being reviewed by Defence for widespread use.
Defence AI projects are being asked to consider ethical considerations such as who is responsible for the AI's decision and actions, and are they lawful, trusted and controlled.
Ensuring the actions of AI are recorded for examination and review will increasingly become important for legal investigations, both on the battlefield and other operations of Defence.
Ethical AI for Defence checklist
- Describe the military context in which the AI will be employed
- Explain the types of decisions supported by the AI
- Explain how the AI integrates with human operators to ensure effectiveness and ethical decision making in the anticipated context of use and countermeasures to protect against potential misuse
- Explain framework/s to be used
- Employ subject matter experts to guide AI development
- Employ appropriate verification and validation techniques to reduce risk.
Ethical AI risk matrix
- Define the activity you are undertaking
- Indicate the ethical facet and topic the activity is intended to address
- Estimate the risk to the project objectives if issue is not addressed
- Define specific actions you will undertake to support the activity
- Provide a timeline for the activity
- Define action and activity outcomes
- Identify the responsible party(ies)
- Provide the status of the activity
For AI projects with the highest levels of ethical risk, the paper recommends a more comprehensive legal and ethical program plan that will help ensure a contract meets the Commonwealth's obligations, including under international weapons agreements.
Stephen Bornstein, who runs Cyborg Dynamics Engineering, has been using the tools while developing his Athena AI technology that has the ability to identify protected objects, people and symbols - such as hospitals, Red Cross or journalists - in near real-time for military operations.
Mr Bornstein said companies working on innovative technology needed to think about ethical considerations not just at a strategic and governance level but also at an engineering level.
"Having an ethical framework to help develop our software and our algorithms and deploy that has huge impact on our entire product lifecycle," Mr Bornstein told The Canberra Times.
Air Force's Air Vice-Marshal Cath Roberts said practical application for the paper's ethical AI framework exists in future human-machine teaming projects as well as current projects like the Loyal Wingman.
"We need to ensure that ethical and legal issues are resolved at the same pace that the technology is developed," Air Vice-Marshal Roberts said.