"... be careful about how you deploy the technology. You can do research, you can do it responsibly, you can publish papers, you can make the public aware of it, but you don't have to give away the code, you don't have to give away the data, you don't have to literally hand a loaded gun to the average person on the internet in order to push the boundaries of science and technology."
- #HanyFarid
https://www.rnz.co.nz/national/programmes/saturday/audio/2018711251/hany-farid-are-deep-fakes-a-real-problem
As long as you're not sharing the software in binary form either, not sharing the source code is ethical from a #SoftwareFreedom POV. But I'm not sure it is from an academic POV. Making claims about research outcomes without sharing enough of your methods to allow them to be independently reproduced strikes me as the PT Barnum school of research.
@strypey as RMS says, software that is not distributed is #freesoftware "in the trivial sense" as long as the 4 freedoms are granted to all those who do receive a copy. It's indeed an academic PoV and not too practical. Apparently just to stress that distribution is not a condition of free s/w.
@strypey as soon as someone somewhere receives a binary w/out source, all they have is trust of those who have eyes on the code which is generally inadequate.