A strong HL example is New York City Local Law 144, a policy intervention regulating automated hiring systems. It addresses potential discrimination by requiring employers to audit certain tools for bias before using them in hiring or promotion.
System: An automated employment decision tool (AEDT) uses machine learning, statistical modelling, data analytics or artificial intelligence to produce a score, classification or recommendation that substantially assists an employment decision.
Example specifics: Under Local Law 144, employers and employment agencies using covered AEDTs in New York City must obtain an independent bias audit conducted within the previous year. They must publish a summary of the results and notify affected candidates or employees before using the tool. Enforcement began in July 2023, so this development postdates the 2022 coursebook.
Impacts and implications: The audit compares selection or scoring rates across sex, race and ethnicity categories, including intersectional categories. This may reveal whether an automated system disadvantages particular groups and gives applicants greater transparency.
The intervention changes the distribution of power by making employers publicly accountable for systems that might otherwise operate as black boxes. It also raises questions of values and ethics, particularly whether numerical fairness measures adequately represent discrimination experienced by individuals and communities.
| Strength | Limitation |
|---|---|
| Independent audits can expose unequal outcomes. | The law does not require employers to correct bias identified by an audit. |
| Public results and notices increase transparency. | It covers only particular hiring and promotion tools used within its jurisdiction. |
| Annual auditing recognizes that systems and data can change. | An audit may measure unequal outcomes without explaining their causes. |
This intervention mainly intercedes in potentially discriminatory hiring processes and may mitigate unequal outcomes, but it does not guarantee fairness.
Exam technique: For an HL Paper 3 response, identify the intervention, explain how its digital system operates, then evaluate it using criteria such as equity, acceptability, feasibility and ethics. Avoid the misconception that completing a bias audit automatically removes discrimination.
