Federal Website Launched to Guide Employers on Avoiding AI Discrimination in Hiring
Washington, D.C., September 25, 2023
The U.S. Department of Labor has launched a new website aimed at assisting employers in mitigating the risks of discrimination through the use of artificial intelligence (AI) in their hiring processes. This initiative, rolled out on September 24, serves as a comprehensive guide offering best practices and a detailed 10-step roadmap for organisations to adhere to laws and maintain fairness.
The guidance is part of broader efforts stemming from a 2023 Executive Order by President Biden, which directed various federal agencies to address urgent AI-related concerns. The Department of Labor's Office of Disability Employment Policy, through its Partnership on Employment & Accessible Technology (PEAT), is spearheading this initiative.
Key Guidelines from the 10-Step Roadmap
The newly launched website offers an array of best practices, which although non-binding, provide a meticulous framework for employers.
1. Identify Legal Requirements:
Employers are urged first to familiarise themselves with pertinent employment non-discrimination, accessibility, and privacy laws. Aligning AI usage with these legal standards is essential to manage risks effectively.
2. Establish Staff Roles:
Organisations should set clear roles for workers involved in AI deployment and provide necessary training. Leadership should be actively engaged, ensuring an organisational culture of accountability and compliance.
3. Inventory Your Technology:
Collect and maintain detailed information about the AI technologies in use, particularly those from third-party vendors. The technology’s intended use, benefits, and risk classification should be thoroughly documented.
4. Work With Vendors:
Engagement with AI vendors is crucial. Employers need to establish consistent policies to ensure vendors adhere to compliance requirements. Identifying and managing risks associated with third-party technologies is a key part of this step.
5. Assess AI’s Impacts:
Conduct thorough assessments of both the positive and negative impacts of AI technology. This includes developing strong risk management policies and collaborating with internal and external stakeholders.
6. Provide Accommodations:
Employers should establish a process allowing job applicants to request accommodations if they feel uncomfortable with AI tools. Although not federally mandated, this step helps to align with broader ADA and state disability laws.
7. Use Explainable AI:
Transparency is emphasized through the use of explainable AI, which involves understanding and documenting the decision-making processes of AI tools. Employers should provide clear, accessible notices to external users describing the technology's operations.
8. Ensure Human Oversight:
Human oversight policies and procedures are necessary for maintaining control over AI processes. This includes training workers on overseeing AI tools and ensuring human accountability remains at the forefront.
9. Manage Incidents:
Organisations should prepare to handle incidents related to AI misuse or failure. Policies for prompt incident management, including regular monitoring and recording of such events, are recommended.
10. Monitor Your AI Regularly:
Continuous monitoring of AI technology performance is crucial. This includes collaborating with various stakeholders to assess reliability and compliance, and swiftly addressing any drift or errors in AI models.
Practical Implications for Employers
The rollout of this website provides employers with a robust framework for incorporating AI in hiring practices while minimizing legal and ethical risks. By adhering to these guidelines, organisations will be better positioned to defend against potential accusations of AI-induced bias and discrimination.
This roadmap recognizes the evolving nature of technology and encourages continuous assessment and improvement. While it does not carry the force of law, compliance with these best practices can offer significant benefits, including enhanced hiring processes and legal safeguarding.
Employers are advised to progressively implement these practices, tailoring their approach based on the specific needs and existing capabilities of their organisations. This iterative implementation allows for more manageable and efficient integration of AI technologies.
This initiative highlights the growing role of AI in the workplace and underscores the importance of regulatory measures to ensure fairness and compliance in emerging technological landscapes.
Source: Noah Wire Services