I’d lean toward option 2, with a little tweaking to fit the practices of my organization. In this case, a scientific justification would be critical to the procedure. Bringing the matter before your Stability Review Board would put the best minds on the case, utilizing their collective expertise and experience.
Given the rise of the AI-induced “too much information” paradigm, now would be a good time to form/engage an AI Utilization Committee (as discussed in the April 9 2024 SH Newsletter article) to decide in advance how these situations will be handled.
