Jisc has outlined a pathway involving four main steps towards the ethical use of artificial intelligence.
The membership organisation technology provider to higher education has published a report on the issue, A pathway towards responsible, ethical AI, saying this could be a better approach that jumping straight into ethics.
The first step is to ask whether a proposal to use AI fits the institution’s objectives. This should be broken down into three discussions on why it is being done, whether it will work and if it can advance the mission.
Second is ‘Does using AI fit our institution’s purpose and culture’, with further questions on what level of ethical complexity is acceptable, does AI fit the local purpose, community and culture, and is it a less intrusive way to perform a process?
Question three is ‘Are we ready to do this?’, reflecting the fact that while some uses of AI in education are in standalone systems others need incorporating into existing systems. This extends to whether the organisation’s supply chain is ready.
Fourthly, there is a question about whether AI raises specific issues, reflecting issues around risk and the differences between artificial and human intelligence. This breaks down into sections around control, explanation, bias and learning.
All this should ease the way into discussions around how a process relates to the law, technology and ethics, the report says.
“The pathway is not meant to involve a full risk assessment, though it may help you decide whether one of those is required,” it adds. “Its discussions may, however, identify risks, mitigations and benefits that are worth noting for later.”
Image from iStock, Monstij