Streamline Local Law 144 compliance with Holistic AI

A bias audit is impartial evaluation of an AEDT by an independent auditor. The aim of a bias audit is to identify and mitigate any biases that may exist in the tool.

The analysis is carried out based on success rates for different demographic groups based on sex and race/ethnicity.

An automated employment decision tool is a computational process that uses machine learning, statistical modeling, data analytics, or artificial intelligence to aid or automate decision-making for employment-related decisions. The tool produces a simplified output, such as a score, classification, or recommendation, that employers can use to screen, evaluate, or select candidates for employment or promotion.

Examples of automated employment decision tools include:

• Resume screening tools that use algorithms to analyze resumes and job applications to identify the most qualified candidates for a particular role.
• Video interviews that are scored using algorithms based on features extracted from candidate responses to assess job-relevant abilities.
• Game-based/image-based assessments that use algorithms to infer relevant abilities based on responses and test-taking behaviours

These tools may result in continuous outputs such as a score or ranking or may result in categorical outputs such as a recommendation on whether or not to progress a candidate to the next stage.

Local Law 144 does not have any exemptions. It applies to all employers or employment agencies using a qualifying AEDT to evaluate employees for promotion or candidates for employment in New York City.

Employers or employment agencies in scope of New York City Local Law 144 can receive penalties for non-compliance. Not having a valid bias audit, not providing a summary of results, and not providing notification all constitute separate violations.