California has introduced numerous proposals aimed at regulating artificial intelligence (AI) this year, addressing concerns such as privacy, bias, and potential catastrophic events caused by errant AI integration into critical systems. The California Civil Rights Department has proposed regulations prohibiting employers from using AI for discriminatory decision-making in hiring processes. The regulations would require internal bias reviews for neural networks and mandate institutions to disclose the data sets used to train their AI.
One real-world example of bias in AI was seen in Amazon’s hiring algorithm, which discriminated against women due to favoring terms predominantly used by men. Detecting discrimination in AI involves identifying pre-existing bias in data sets or analyzing network outputs for persistent unfavorable outcomes for certain groups. However, regulations may overlook the distinction between problematic and non-problematic bias, potentially subjecting developers and companies to penalties unjustly.
Additionally, Senate Bill 926 in California targets deepfakes by making it illegal to create and distribute AI-generated sexually explicit images misrepresented as authentic. Despite the intent to curb harmful deepfakes, enforcement may prove challenging due to the widespread availability of open-source AI and the anonymity of creators.