The reported complicity of the American Psychological Association (APA) in the 'enhanced interrogation' programme of the US Department of Defense (DOD) represents improper exploitation by the DOD of its relationship with the APA in pursuit of its own objectives, distorting scientific ethics and practices in the process (see Nature 523, 255; 2015). The scientific community as a whole could learn from the mechanisms that the DOD used to influence the APA, as reported in an independent review (go.nature.com/4vpdob).

Historically, similarly inappropriate relationships have sometimes developed between granting agencies and scientists (see, for example, S. Reverby Examining Tuskegee (Univ. North Carolina Press, 2009) and Nature 467, 645; 2010). Scientists' alignment with institutional preferences is often a force for good, but vigilance is needed to prevent erosion of ethical or evidential standards.

Institutional coercion such as that revealed in the report is disastrous for the public perception and integrity of science. The moral force of research standards is weakened if they are subject to political concerns.

We all need to address this potential for abuse without impeding the normal functioning of scientists and their granting institutions. For example, negotiations on ethics and research practices between institutions could be overseen to ensure good-faith participation and eliminate any back-door dealing. Such regulatory authority could be absorbed by existing bodies such as the US Federal Inspectors General.