New Laws Targeting AI Bias in Hiring

Arbor Team

Artificial intelligence has exploded in popularity and adoption over the past decade, and it’s become a major topic of conversation with the recent advent of ChatGPT. More than ever, technologies like machine learning, neural networks, and natural language processing are now being used across industries, from finance to healthcare to transportation. And now, applications of AI are increasingly making their way into human resources and hiring.

As AI becomes ubiquitous in hiring, companies must ensure these technologies are fair, transparent, and accountable. New York City is one of the first to take legal steps toward ensuring responsible use of AI in hiring. In this article, we’ll review some concerns around AI, and the steps that New York is taking in an attempt to minimize unfair outcomes.

AI in hiring

Technologies like natural language processing, computer vision, and predictive analytics are now being widely deployed by companies to screen resumes, conduct video interviews, and even make final hiring decisions.

According to a recent Society for Human Resource Management survey, nearly 1 in 4 employers now use some form of AI technology in their hiring and recruiting efforts. The adoption of AI tools has steadily increased over the past decade as these technologies have become more sophisticated and accessible.

Concerns around bias in AI

While AI-based hiring tools have proven efficient in automating the recruiting process, lack of transparency in these tools have led to concerns around bias in hiring. These concerns stem from various factors, including:

  • Historical biases: AI algorithms are often trained on historical data that may reflect past discriminatory practices against certain demographic groups. The models can perpetuate these biases.
  • Lack of transparency: The "black box" nature of complex AI models makes it difficult to audit for discrimination.
  • Limited diversity in tech: Development teams lacking diversity and awareness of bias issues may not proactively address these problems in algorithm design.
  • Poorly constructed training data: Flawed data collection and sample selection practices when training models can further entrench biases.
  • Built-in subjectivity: Attempts to automate inherently subjective processes like evaluating skills and aptitudes are prone to reinforcing stereotypes.
  • Regulation gaps: New applications of AI often outpace existing anti-discrimination laws, allowing potentially problematic tools to spread unchecked.

Emerging US laws regulating AI-powered hiring tools

These concerns around bias in AI tools have resulted in new legislation to protect candidates from discrimination. The first of which took effect earlier this month with New York City’s regulation of automated employment decision tools (AEDT). Under this law, NYC employers who utilized automated hiring systems must:

  • Provide candidates with a written notice if an automated tool was used to screen or evaluate them
  • Obtain written consents from candidates before utilizing an automated tool and enable candidates to request info on the tool’s determinations and data collected
  • Audit these systems annually for bias and make the audit publicly available

This legislation is enforced by the NYC Commission on Human Rights, and violations can result in civil penalties up to $500,000. Beyond New York, legislators in several other states, including California, Illinois, New Jersey, Washington, and Oregon, have already drafted similar legislation.

Preventing bias in the hiring process

With growing regulation to protect against workplace discrimination, employers should take active steps to ensure compliance and fair hiring practices. Measuring diversity across hiring funnels can help companies stay ahead of potential bias – tracking pipeline diversity across stages and trends over time enables employers to identify potential areas for improvement and help audit the effects of automated hiring tools. Leveraging additional software tools may be helpful for analyzing these trends and ensuring proper consents with candidates.

Adopting AI in the hiring process can be extremely powerful, but it’s important to be proactive in preventing potential bias or discrimination when attracting and evaluating talent.

About Arbor

Arbor enables leaders to easily capture, analyze, and benchmark DEI outcomes. Through integrations into existing HR and applicant tracking systems, Arbor’s dashboard helps employers easily visualize workforce and hiring trends to ensure fair practices and identify potential biases.

Contact us at to learn more about how our platform can help your organization ensure fair hiring and stay on top ahead of new regulations.

Ready to learn more?

We’re working with industry leaders to enable fairer, better workplaces.