BETTER SOCIAL MEDIA FEEDS ACT
Illinois SB3454, introduced February 5, 2026 by Sen. Sue Rezin, creates the Better Social Media Feeds Act (Section 1), effective January 1, 2027 (Section 99). The bill targets covered online platforms that use algorithmic recommender systems and conduct business in Illinois. Under Section 20 (Design Transparency), platforms must prominently disclose on their website or app: (1) a list of each algorithmic recommender system in use, (2) a description of each system's inputs and data sources, and (3) the weights used in each system categorized into quartile groups by relative importance. Platforms must also annually disclose the high-level objectives, key results, and performance metrics used to evaluate product teams responsible for recommender system design. Under Section 25 (User Choice and Defaults), all algorithmic recommender systems using personal data must be configured by default to maximize one or more long-term user value metrics, and platforms must provide an accessible user interface for users to set content preferences. Platforms may not penalize users — through degraded service, lower quality, or higher prices — for exercising these rights. Section 30 extends the same long-term user value default requirement specifically to covered minors. Section 35 requires platforms to maintain at least one holdout group for at least 12 months and conduct annual long-term holdout assessments, publicly disclosing aggregate anonymized metrics comparing the holdout group to the general user base. Independent audits of these assessments are required annually at the platform's expense. The Attorney General is empowered to adopt clarifying rules and, by January 1, 2028, rules governing holdout assessment operations. Under Section 40, violations constitute unlawful practices under the Consumer Fraud and Deceptive Business Practices Act (815 ILCS 505/2MMMMM new), with all associated remedies available to the Attorney General.
Status history
Current status as of 2026-02-05
In committee
2026-02-05
observed 2026-09-02
Impact areas
- Quality Assurance
- AI Policy
- Privacy & Data Protection
- Algorithms & Automated Decisions