Introduction
Recent developments in federal AI policy, particularly the withdrawal of guidance from the Equal Employment Opportunity Commission (EEOC) and the Department of Labor (DOL) concerning AI and workplace discrimination [1] [5], have sparked significant concerns among employers. This shift in policy underscores the importance of understanding and complying with existing anti-discrimination laws and adapting to emerging state-level regulations.
Description
Recent changes in federal AI policy [1] [5], including the rescission of guidance from the Equal Employment Opportunity Commission (EEOC) and the Department of Labor (DOL) regarding AI and workplace discrimination [1] [5], have raised significant concerns for employers. This rollback follows an executive order that aimed to develop security standards and protections for workers, assess AI’s impact on the workforce [2], prevent discrimination [1] [2] [5], and support upskilling programs [2]. The previous order also sought to establish best practices for job displacement and workplace discrimination while facilitating the hiring of AI professionals and streamlining immigration for skilled workers [2]. However, the current federal actions have primarily focused on awareness and guidance [2], leading to increased scrutiny on state-level regulations addressing AI-driven discrimination.
Employers must remain vigilant [1] [5], as the removal of government guidance does not alter fundamental anti-discrimination laws such as Title VII of the Civil Rights Act and the Americans with Disabilities Act (ADA) [1] [5]. Discrimination in hiring processes [4], whether through AI tools or human decisions [4], remains illegal [4], and employees and job applicants retain the right to sue over discriminatory AI practices [1] [5]. Additionally, state and local governments are enacting their own regulations to address algorithmic discrimination, which involves differential treatment based on protected characteristics [4].
Key considerations for employers include:
- Disparate Impact Discrimination: AI tools that unintentionally exclude or disadvantage protected groups may violate Title VII [1] [5]. The effectiveness of AI systems is heavily dependent on the quality of the training data; biased training data can lead to biased outcomes [4].
- Disability Discrimination: AI systems that screen out candidates based on disability-related characteristics could lead to liability under the ADA, necessitating reasonable accommodations [1] [5].
- Vendor Liability: Employers can be held accountable for discrimination resulting from AI tools developed by third-party vendors.
To mitigate bias [4], it is essential for employers to use well-curated data and conduct thorough analyses and tests for biased results [4]. Regular monitoring of AI systems is crucial for compliance with anti-discrimination laws [1] [5]. Employers should conduct internal audits [1] [5], require transparency from vendors regarding AI algorithms [1] [5], and review AI liability provisions in vendor agreements [1] [5]. Additionally, conducting risk assessments for AI-driven employment decisions and implementing transparency measures [3], such as public disclosures and adverse action notices [3], are vital steps to ensure compliance with evolving regulations.
Specific state regulations include:
- Colorado AI Act (effective February 1, 2026): Mandates reasonable care to avoid algorithmic discrimination [1] [5], annual impact assessments [1] [5], and employee notification regarding AI use in consequential decisions [1] [5].
- Illinois HB 3773 (effective January 1, 2026): Amends the Illinois Human Rights Act to prohibit discriminatory AI use and requires notification of AI use in employment processes.
- Illinois AI Video Interview Act (effective January 2020): Regulates AI analysis of video interviews, requiring disclosure [1] [5], consent [1] [5], and deletion rights [1] [5].
- New York City Local Law 144 (effective July 2023): Requires advance notice of automated decision tools, annual independent bias audits [1] [5], and publication of audit results [1] [5].
- California Civil Rights Department Proposals: Aims to prevent the use of AI that screens out applicants based on irrelevant factors, requiring employers to maintain records and conduct anti-bias testing [4].
- Texas Responsible AI Governance Act: Establishes obligations for high-risk AI systems, including risk assessments and transparency measures [4], with penalties for non-compliance [4].
Employers must navigate a complex landscape of state and local AI regulations [1], especially as numerous bills are being introduced to regulate AI in employment [1]. Best practices include:
- Implementing an AI use policy [1] [5].
- Conducting regular AI audits to identify potential biases.
- Reviewing vendor agreements for compliance and transparency.
- Ensuring human oversight in AI decision-making processes, which is critical for compliance with emerging regulations [2].
- Monitoring legal developments to adapt proactively to new regulations.
The removal of federal AI guidance does not diminish the legal risks associated with AI in employment decisions [1]. Employers are advised to remain informed about relevant legislation and prioritize transparency and proactive audits to manage bias risks in AI tools [4]. Establishing compliance strategies now is essential to avoid penalties as regulations, such as those in Colorado, come into effect in 2026.
Conclusion
The withdrawal of federal guidance on AI and workplace discrimination necessitates heightened awareness and compliance with existing anti-discrimination laws. Employers must proactively adapt to a complex regulatory environment, particularly at the state level, to mitigate legal risks. By implementing robust compliance strategies, conducting regular audits [1] [5], and ensuring transparency, employers can effectively manage the challenges posed by AI in employment decisions and avoid potential penalties as new regulations come into force.
References
[1] https://www.jdsupra.com/legalnews/thought-leadership-ai-and-workplace-1405460/
[2] https://www.shrm.org/topics-tools/news/technology/trump-issues-executive-order-to-boost-ai
[3] https://blog.cisive.com/jan-2025-employment-law-updates
[4] https://www.salary.com/newsletters/law-review/state-regulation-of-ai-where-we-are-now/
[5] https://www.huschblackwell.com/newsandinsights/ai-and-workplace-discrimination-what-employers-need-to-know-after-the-eeoc-and-dol-rollbacks
