New York City Law Aimed at Curbing AI Bias Goes Into Effect

New legislation focused on artificial intelligence-driven employment tools went into effect in New York City on Wednesday, prohibiting employers and agencies from using automated employment decision tools (AEDT) unless that tool was audited for bias within the last year.

Based on the 2021 Local Law 144, the new rule aims to prevent bias and ensure fairness in employment decisions using artificial intelligence.

In addition to the one-year audit window, the NYC Department of Consumer and Worker Protection (DCWP) rule also requires the audit results to be available to the public, plus employees and job candidates must be provided appropriate notices.

An automated employment decision tool assists or replaces discretionary decision-making. Examples of these tools include Freshworks, Breezy HR, Greenhouse, and Zoho Recruits, which screen or pre-screen resumes.

“DCWP will collect and investigate complaints [of bias],” a department spokesperson told Decrypt via email. “Calculations of civil penalties for violations will be case specific,” the agency added, explaining that the law allows for separate violations for each day that AEDT usage violates the law.

In June, a business review website B2B Reviews report said 21% of HR departments surveyed used AI tools like OpenAI’s ChatGPT for training and development, saving on average 70 minutes per week. One in 10 HR departments also said they used the chatbot to draft termination letters.

“In developing the AEDT rule, we worked to strike the appropriate regulatory balance between the rights of job applicants and the needs of businesses, as reflected in our rulemaking process,” the spokesperson said.