MIT researchers have been able to assign cafeteria bills to individuals with 80% probability. With any identification information as well as timestamps removed.

MIT researchers have been able to assign cafeteria bills to individuals with 80% probability. With any identification information as well as timestamps removed. Correlating datasets gets pretty powerful pretty fast. Also, if the data is made public, we are basically betting, that it is not possible to deanonymize in the future. Given the rate of technological development, I’d be very reluctant to make that bet.