[smaudet]

> I don't see how regulating algorithms and AI could affect those institutional injustices.

Because this is a dangerous technology which has only proven, again and again, to be unreliable, unethical, inaccurate, and is EASILY used by the people most likely to abuse it.

The point is not that the dutch tax office is corrupt, it is that corrupt individuals and entities are likely, and have used, this for nefarious, foul purposes. Regulating and/or banning its use is only one of many avenues to attempt to safeguard (against) this tech.

[eesmith]

In my reading, people aren't saying the tax office corrupt, but rather that it was illegally discriminatory. Eg:

> the Dutch data protection agency also fined the Dutch tax administration €2.75 million in December 2021 for the “unlawful, discriminatory and therefore improper manner”

My point is that AI regulation is insufficient, not that it's unnecessary. Eg, the Amnesty International report says:

> civil servants spoke in denigrating terms about families with Caribbean roots, referring to them as an “Antillean nest”. Civil servants flagged these applicants through manual selection.

Manual selection is independent of AI.