Unfortunately AI models like this one often never make it to the clinic. The model could be impressive enough to identify 100% of cases that will develop breast cancer. However if it has a false positive rate of say 5% it’s use may actually create more harm than it intends to prevent.
Personally I’m not sure the gate keeping you’re observing is all that much of an issue. I think it’s useful to remember many vegans are also public advocates for veganism. It’s important to them that people generally know what they mean when they advocate for veganism.
However, the definition of all words are always in flux. It’s not uncommon to see people call themselves vegan when a more apt description of their lifestyle would be plant based, flexitarian, vegetarian, etc. As such, I think edge cases like your friend take on an outsized importance that goes beyond the morality of your friend eating honey.
Basically, the goal may not be the social exclusion of your friend which is what I think is usually the problematic aspect of gatekeeping.