> “It could be unfair to do this, to sort of induce people to rely on the email alerts,” Bruckner explained, saying that if Coinbase started and stopped email a

> “It could be unfair to do this, to sort of induce people to rely on the email alerts,” Bruckner explained, saying that if Coinbase started and stopped email alerts without properly notifying users and that this was proven to have “caused substantial injury” like financial losses, that could present an issue. It seems pretty ridiculous to me to assert that Coinbase stopping email price alerts is liable for investment losses, but the professor added the appropriate caveats, so my criticism is mostly with the publisher, not the source. The interesting aspect to this, to me, is the idea that a/b testing could interfere with legal obligations (the article implies but does not definitively state that a/b testing was involved). I've worked on a/b test systems in ecommerce, and adding those obligations would have meant enough new requirements to substantively change the resulting designs. This would be an interesting question to have to answer: If this test has a statistically significant result, would that mean one side of the population experienced better financial results than the other, caused by our actions?