Transparency of algorithmic decision-making: Limits posed by IPRs and trade secrets

Today, algorithms are assuming a leading role in various areas of decision-making. Prompted by a promise to provide increased economic efficiency and fuel solutions for pressing societal challenges, algorithmic decision-making (ADM) is often celebrated as an impartial and constructive substitute for human adjudication. But in the face of this implied objectivity and efficiency, the application of algorithms is also marred with mounting concerns about embedded biases, discrimination, and exclusion. Transparency as a figure of accountability and a measure of compliance needs to be effective. As this contribution shows, however, uncertain contours of IP and trade secret protection might frustrate this goal. Through examples of several data subjects’ rights, it demonstrates that trade secrets in particular pose a particular challenge to the proposed transparency. The ADM context has drawn much attention owing to the great consequential impact it has on our everyday lives. However, it seems to merely amplify the embedded challenge of defining the scope of trade secrecy protection in principle. Whereas the CJEU is expected to offer some guidance with regard to Article 22 and the extent of the information to be disclosed, it appears that a thorny question of trade secrets and data access extends much beyond.