"As we've said before, the fact that states are regulating AI is often a good thing. Left unchecked, company and government use of automated decision-making systems in areas such as housing, health care, law enforcement, and employment have already caused discriminatory outcomes based on gender, race, and other protected statuses. While state AI laws have not been perfect, they are genuine attempts to address harms that people across the country face from certain uses of AI systems right now."
"For example, a copy of the draft order published by Politico specifically names the Colorado AI Act as an example of supposedly "onerous" legislation. As we said in our analysis of Colorado's law, it is a limited but crucial step-one that needs to be strengthened to protect people more meaningfully from AI harms. It is possible to guard against harms and support innovation and expression."
A draft executive order would direct federal agencies to legally challenge state AI laws deemed onerous, restrict federal funding to states with such laws, and seek federal legislation to override state AI regulations. State-level regulation often aims to address present harms from automated decision-making in housing, health care, law enforcement, and employment that have produced discriminatory outcomes for protected groups. State laws like Colorado's represent limited but important initial protections that require strengthening. Federal preemption risks halting state progress, weakening safeguards, and leaving people exposed to discriminatory AI-driven decisions.
Read at Electronic Frontier Foundation
Unable to calculate read time
Collection
[
|
...
]