A Deployment Model to Extend Ethically Aligned AI Implementation Method ECCOLA

AI ethics remains disconnected from practice: The paper identifies a critical gap where AI ethics discussions stay largely theoretical while technical teams developing AI systems lack practical tools to implement ethical considerations, meaning good intentions often don't translate into ethical outcomes.

ECCOLA method shows promise but needs adoption support: While the ECCOLA framework has demonstrated success in increasing ethical considerations during AI development, the paper emphasizes that even well-designed methods fail without actual use—highlighting that adoption and implementation are as important as the method itself.

New deployment model introduces measurable ethics metrics: The extended framework provides "simple metrics to facilitate the communication of ethical gaps," offering organizations concrete ways to measure and discuss ethical outcomes rather than relying on vague concepts, making ethics discussions more tangible for technical teams.

Ethics assessment can happen at any stage: The model enables evaluation of AI systems' ethical alignment at different points in their lifecycle—including when acquiring existing AI systems—rather than only during initial development, expanding opportunities to catch and address ethical issues.

Bridge between ethics and technical practice is essential: The paper underscores that practitioners in software development need tools specifically designed to help them process ethical considerations alongside technical tasks, not separate from them, suggesting ethics integration requires rethinking how development teams work.