New Trend: “Algorithmic Discrimination” in AI
The rise of artificial intelligence (AI) is transforming the workplace in unprecedented ways, especially in the automation of the hiring process. If you’ve applied for jobs recently, you know just how frustrating it can be to navigate this process. Not surprisingly, as AI becomes more integrated into employment practices, it brings to the forefront critical questions not only about accuracy, but also about fairness, discrimination, and privacy.
There is talk of a new trend called "algorithmic discrimination"— where biases within AI systems lead to unfair treatment in hiring, promotions and evaluations. As you know, AI has to be “trained” over time to be able to provide accurate results. Without a large enough dataset for the training, the results can prove problematic.
Although there are no current federal laws to tackle holding employers accountable for biased outcomes generated by AI, we look for this to be a growing area of legal focus.
Several states, and New York City, have already enacted laws against algorithmic discrimination. For example, Colorado, Delaware and Illinois have enacted legislation addressing this issue, while New York state and New Jersey currently have bills pending.
- New Jersey’s law addresses the requirement for an independent bias audit of AI tools in screening candidates and other employment decisions.
- New York’s proposed legislation also requires an independent bias audit and focuses on increasing transparency and accountability for AI systems used during the hiring and promotion processes.
- Delaware’s recently passed legislation aims to ensure transparency and fairness in the use of AI for employment decisions, particularly in hiring practices. More specifically, this legislation requires employers to provide clear disclosures about how AI systems are used and to take measures to prevent discrimination.
No legislation is currently pending in Pennsylvania. We’ll keep an eye out for any additional developments in this area and share updates accordingly. You can read more on recent legislative developments regarding “algorithmic discrimination” here.