The European Commission said that companies engaged in artificial intelligence need to establish accountability mechanisms to prevent misuse of technology under new ethical guidelines for a technology open to “authoritarian regimes” and that artificial intelligence projects must be transparent, human-supervised and logarithms safe and reliable, subject to privacy and data protection rules, as well as other recommendations.

The new EU initiative is trying to intervene in a global debate about when or whether companies should make moral concerns to commercial interests, and how regulators can afford new projects without risking killing innovation.

Last year, the EU’s chief executive helped 52 academic, industrial and corporate experts including Google, SAP, Santander and Bayer to help formulate principles.