Report prepared by the Montreal AI Ethics Institute (MAIEI) on Publication Norms for Responsible AI

Publication decisions need better guidance tools: The report recommends creating tools to help researchers navigate ethical publication decisions, recognizing that the fragmented nature of AI research makes it difficult for scientists to know what responsible disclosure looks like across different contexts.

Impact statements should be mandatory and comprehensive: MAIEI calls for "broad impact statements" in AI research publications, requiring authors to consider potential negative consequences and real-world applications of their work—moving beyond traditional abstract summaries to address societal implications.

Peer review processes must evolve: The report identifies peer review as needing fundamental changes to adequately evaluate responsible AI research, suggesting current review systems aren't equipped to assess ethical dimensions and potential harms alongside technical merit.

Expected results should be published, not just successful outcomes: Requiring researchers to publish expected results alongside actual findings would help prevent selective reporting and give the scientific community a more complete picture of what AI research actually achieves.

Balancing responsibility with innovation concerns: The report acknowledges legitimate concerns that stricter publication norms could constrain innovation or push sensitive research underground, indicating that responsible AI governance requires careful calibration rather than heavy-handed restrictions.