Purpose

This bill aims to stop using surveillance-based methods to set prices for goods or services and wages for workers. It would prohibit automated decision systems that tailor prices or pay to individuals based on surveillance data, and it would require transparency and protections for consumers and workers.

Key definitions (to understand how the bill works)

• Automated decision system: a system or process that uses computation (including machine learning, statistics, or AI) to help or replace human decisions.
• Surveillance data: data gathered by observing or collecting information about a person, including personal characteristics, behaviors, biometrics, or related inferences (e.g., browsing history, IP address, location, purchase history).
• Behaviors: observable or inferred actions, habits, preferences, or vulnerabilities of a person.
• Biometrics: unique biological or behavioral traits used to identify someone (e.g., fingerprint, voice, facial features, iris, genetic information).
• Personal characteristics: traits like race or eye color, as well as mutable details like address, weight, citizenship, or parenthood, and other identifying information.
• Price: the total amount charged for a good or service, including related fees and terms.
• Wage: all terms of pay for work (hourly, salary, bonuses, incentives, scheduling, task assignment, etc.).
• Consumer: a person buying goods for personal or household use.
• Worker: a person who works for an employer (employee, independent contractor, etc.).
• Insurer: an insurance company.
• Surveillance-based price discrimination: using an automated decision system to set individualized prices based on surveillance data about a consumer.
• Surveillance-based wage discrimination: using an automated decision system to set individualized wages based on surveillance data about a worker.
• Differential pricing: different prices charged to different people.
• Consumer report (FCRA): information used in credit decisions that is regulated by the federal Fair Credit Reporting Act.

Main provisions

• Subd.2. Price discrimination
  • Prohibition: It is illegal to engage in surveillance-based price discrimination.
  • Allowed (non-discriminatory) pricing under certain conditions: 1) Prices that differ are justified by real cost differences in providing the good or service. 2) Discounts offered to all consumers on equal terms, if:
    • the discount terms are publicly available, and
    • the discount rewards membership in a particular group (e.g., active military, veterans, teachers, students, seniors). 3) If you are an insurer and you only use risk-relevant data to determine insurance pricing.
  • Other exception:
  • A business is not considered to engage in surveillance-based price discrimination if it does not extend credit under terms based on a consumer report covered by the Fair Credit Reporting Act.
• Subd.3. Wage discrimination
  • Prohibition: It is illegal to engage in surveillance-based wage discrimination.
  • Allowed (non-discriminatory) wage decisions under certain conditions: 1) Wages are individualized based only on data specific to the worker that directly relates to the tasks the worker was hired to perform, or 2) Differences in the cost of providing the labor to the employer. 3) Before hiring, the employer must plainly disclose to all workers what data is considered and how the automated decision system uses it.
  • Note: Not hiring a person who has not previously worked for the employer or its affiliates/subsidiaries is not itself considered wage discrimination under this bill.
• Subd.4. Publication of procedures
  • If an automated decision system is used to help set wages or prices, the employer must publish reasonable procedures to: 1) Ensure the accuracy of all data considered by the system, 2) Allow consumers or workers to correct or challenge the accuracy of data used by the system, 3) Inform consumers or workers what data is considered and how the data affects price or wage decisions.

Significant changes this bill would make

• Creates a statewide prohibition on surveillance-based price and wage discrimination using automated decision systems.
• Establishes defined terms for automated decision making, surveillance data, and related concepts, bringing clarity to how such systems could affect pricing and pay.
• Introduces strict transparency and accountability requirements, requiring public-facing procedures to ensure data accuracy and to inform and empower consumers and workers to challenge or understand data used in pricing or wages.
• Sets specific exceptions to allow cost-based pricing, publicly disclosed discounts, group-related discounts, and insurer use of risk-relevant data, while preserving protections against discriminatory practices.
• Requires pre-employment disclosures about data and how it is used, linking to fair hiring practices.

How this would apply

• Companies and insurers would need to assess their pricing and wage practices to ensure they do not rely on surveillance data to tailor prices or pay.
• If using automated decision systems, they would need to document and publish procedures for data accuracy, correction rights, and disclosure of data inputs and decision logic.
• Certain legitimate pricing/wage practices based on cost, public discounts, or special groups could continue if they meet the specified criteria and disclosures.

---

Relevant Terms - surveillance-based price discrimination - surveillance-based wage discrimination - automated decision system - surveillance data - price - wage - consumer - worker - insurer - personal characteristics - biometrics - cost to provide - discounts - group membership (military, veterans, teachers, students, seniors) - FCRA (Fair Credit Reporting Act) - consumer report - data accuracy - correction/challenge - disclosure of data used in decisions - plain language disclosure - pre-hire disclosure