AI unmasks anonymous chess players, posing privacy risks

Re:AI unmasks anonymous chess players, posing privacy risks by dewey on Sunday December 11, 2022 @09:13AM (#33943346)

Basic opsec principles always say that if you want to stay anonymous you have to switch the way you are writing (by adopting a differnet personality, running it through translation apps etc.).

If someone would want to stay anonymous the unmasking % would probably be lower. The threat-model of the chess players doesn't include that they have to stay anonymous and need to switch up their way of playing.

[ Parent | on HN ]

Re:AI unmasks anonymous chess players, posing privacy risks by greggarious on Sunday December 11, 2022 @11:47AM (#33944785)

I’m fond of occasionally throwing in some ou s and references to cities other than Ontario to throw folks off, but it’s hard to pull off long term

[ Parent | on HN ]

Re:AI unmasks anonymous chess players, posing privacy risks by jojo259 on Sunday December 11, 2022 @07:28PM (#33949294)

Just realized the fact I type like an American online (I'm British) would confuse someone for a little bit if they tried to de-anonymize me... (although they probably wouldn't struggle anyway)

[ Parent | on HN ]

Re:AI unmasks anonymous chess players, posing privacy risks by Havoc on Sunday December 11, 2022 @04:12PM (#33947579)

>>switch the way you are writing

I'd be very surprised if that actually works. Stuff like vocabulary can't exactly be turned off at will

[ Parent | on HN ]

Re:AI unmasks anonymous chess players, posing privacy risks by dewey on Sunday December 11, 2022 @06:06PM (#33948616)

It probably got a lot easier with ChatGPT now.

[ Parent | on HN ]