Hi Anna,

My immediate response is to leave the current default as it is – i.e. no AI scraping/training of member articles.

The alternative (individual opt-in/opt-out) would be a monster to manage (let alone define), and most certainly would not suit across the spectrum of members – the proverbial can-of-worms!

Further, what control would we have over the use being put to the outputs of these “trained” systems? Conflicts of values could easily arise (and could actually damage reputation).

Further still, what actual additional exposure would opting-in achieve? A few more external views, for example, would have limited value.

So for what it's worth, and what I am worth,and given my background in AI is somewhat meagre, this is my offering for now.

Hope it helps.