New York City's Local Law 144 requires employers and employment agencies that use automated employment decision tools (AEDTs) to obtain an independent, impartial bias audit annually. The law applies to any computational process derived from machine learning, statistical modelling, data analytics, or artificial intelligence that results in a simplified output used to substantially assist or replace discretionary decision making in employment decisions. Employers must identify if their tools fall under the regulations, identify an independent auditor to conduct the audit, prepare the required data for the audit, and establish notification procedures. Failure to comply could result in limited access to a talent pool in New York City.

California policymakers are proposing modifications to employment regulations to ensure that automated decision systems (ADSs), which are being increasingly used for talent management, do not result in unlawful discrimination based on protected characteristics. ADSs include algorithms used for screening resumes and online tests to measure skills, dexterity, reaction time, and cultural fit. The proposed modifications define ADSs, algorithms, artificial intelligence, machine learning, adverse impact, and proxy. They require evidence of bias testing and retention of records, including ADS data for at least four years. Employers must ensure ADSs measure job-related characteristics and are appropriately validated to avoid unintentional discrimination.

The Digital Services Act is a new EU law that regulates digital services and requires companies to assess risks, outline mitigation efforts, and undergo third-party audits for compliance. The Act applies to hosting services, marketplaces, and online platforms that offer services in the EU, regardless of their place of establishment. Key provisions of the Act include transparency and algorithmic accountability and a risk governance approach to AI systems. Non-compliance with the Act can attract penalties of up to 6% of the company's annual turnover in the European Union and companies will also be exposed to civil suits and liability. The Act will be in full effect starting February 17th, 2024.

The U.S. Equal Employment Opportunity Commission (EEOC) has released a "technical assistance document" advising employers on how to assess adverse impact in software, algorithms, and artificial intelligence used in employment selection procedures under Title VII of the Civil Rights Act of 1964. The document provides key definitions for software, application software, algorithm, and artificial intelligence, and emphasizes that employers are accountable for any disproportionate impacts caused by AI selection tools. The document also encourages employers to regularly self-audit selection tools to assess any adverse impact on protected groups and modify the tool to minimize such effects. The EEOC's stance reaffirms the agency's position that AI systems and tools must adhere to laws and regulations concerning equal employment opportunity.

New York City's Bias Audit Law, or Local Law 144, which comes into force on July 5th, 2023, requires that employers and employment agencies using automated employment decision tools (AEDTs) to evaluate candidates for employment or employees for promotion have undertaken a bias audit of the tools and announced procedures for notification. However, the definition of key elements of the legislation has evolved throughout the rulemaking process, with the final version providing clarification on definitions for machine learning, statistical modeling, data analytics, or artificial intelligence, simplified outputs, substantially assisting systems, independent auditors, calculating impact ratios, summary of results requirements, and using historical and test data. While the adoption of AI and automation in HR is becoming ubiquitous, concerns about existing biases being perpetuated and amplified, novel sources of bias, and a lack of transparency about the system's capabilities and limitations remain.