Data & Society AI on the Ground Program Director Jacob Metcalf hosts a conversation with experts on the emerging regulations in the EU, US, and Canada, and asks how impact assessment practices can effectively account for the harms caused by algorithmic systems.
Algorithmic impact assessments have emerged as a centerpiece of the conversation about algorithmic governance. Impact assessments integrate many of the chief tools of algorithmic governance (e.g., auditing, end-to-end governance frameworks, ethics reviews) and speak to the challenges of algorithmic justice, equity, and community redress. Impact assessment (or similar accountability mechanisms) are at the core of recent headlines about procurement practices in Canada, leaked regulatory proposals in the EU, and new efforts to regulate the tech industry in the US. But do impact assessments promise too much? Multiple national and state governments have instituted, or are considering, requirements for impact assessment of algorithmic systems, but there is a surprisingly wide range of structures for these regulations.