The New Jersey Division on Civil Rights (DCR) recently released guidance on how the New Jersey Law Against Discrimination (LAD) applies to algorithmic discrimination. This guidance clarifies that the LAD prohibits algorithmic discrimination in various settings, including employment, housing, and places of public accommodation.
Background
Automated decision-making tools, powered by artificial intelligence (AI) and algorithms, are increasingly used in various sectors. While these tools offer potential benefits, they also carry the risk of algorithmic discrimination.
Understanding Algorithmic Discrimination
Algorithmic discrimination refers to discrimination resulting from using automated decision-making tools. These tools can perpetuate bias and lead to discriminatory outcomes if not designed and used responsibly.
How the LAD Protects Against Algorithmic Discrimination
The LAD prohibits discrimination based on various protected characteristics, including race, religion, color, national origin, sexual orientation, gender identity, and disability. This protection extends to algorithmic discrimination, regardless of whether it results from disparate treatment, disparate impact, or failure to provide reasonable accommodations.
Key Takeaways for Employers
- Employer Liability: Employers and other covered entities are liable for any discrimination resulting from their use of automated decision-making tools, even if the tools were developed by a third party.
- Mitigating Risk: To mitigate the risk of algorithmic discrimination, employers should carefully evaluate the design and testing of automated tools before and after deployment. This includes implementing quality control measures for data used in automated tools, conducting impact assessments and bias audits, and providing notice of the use of automated decision-making tools.
Next Steps for Employers
To ensure compliance with the LAD’ guidance and minimize the risk of algorithmic discrimination, employers should:
- Stay Updated: Review the full DCR guidance on algorithmic discrimination.
- Conduct an Assessment: Conduct a self-audit of existing automated decision-making tools to assess potential bias.
- Educate and Train: Train HR staff and other relevant personnel on algorithmic discrimination and its implications.
- Review Policies: Consider implementing a policy addressing the use of automated tools in employment decisions.
- Seek Guidance: Consult with legal counsel to ensure compliance with the LAD.
