Non-technical Methods of Achieving Trustworthy AI
Dominika Ewa Harasimiuk, Tomasz Braun · 2021
While for developers and deployers of AI technologies crucial are technical methods enabling the achievement of trust in AI, for the lawyer, special attention shall be drawn to non-technical means. Regulatory approaches may take the form of the risk-based approach. In the areas where there is a substantial threat of harm, the risk-based regulatory approach shall be rather replaced by the precautionary principle-based one. The intensity of regulation shall depend on the level of autonomy in AI-based decision making. Not only statutory measures shall be adopted, but stakeholders shall adapt their corporate structures. Also, standardisation and certification are one of the non-technical methods having potential in the AI field. Finally, inclusiveness and education — informed participation, social dialogue, increasing AI literacy are the goals for the upcoming years.