Regulon.AI — NY-A-06974 · NY A 6974 (SAFE for All Act)

Summary

The SAFE for All Act requires operators of addictive social media platforms to provide users with mechanisms to turn off algorithmic recommendations, disable notifications (including a nighttime-only option), turn off autoplay, and set daily time limits on platform access. Operators are prohibited from using dark patterns to subvert these user controls or to make account deactivation or cancellation more difficult. Enforcement is exclusively through the New York Attorney General, who may seek civil penalties up to $5,000 per violation, injunctive relief, restitution, disgorgement, and damages. The bill expressly preserves the SAFE for Kids Act and COPPA requirements.

Enforcement & Penalties

Enforcement Authority

Exclusive enforcement by the New York Attorney General, who may bring an action or special proceeding upon complaint or otherwise. The AG must maintain a public website to receive complaints, information, and referrals from the public. No private right of action is created.

Penalties

Civil penalties of up to $5,000 per violation. The AG may also obtain injunctive relief (including preliminary relief), restitution of moneys or property obtained by the violation, disgorgement of profits or gains, actual damages caused directly or indirectly by the violation, and any other relief the court deems proper.

Who Is Covered

"Operator" shall mean any person, business, or other legal entity, who operates or provides an addictive social media platform.

What Is Covered

"Addictive social media platform" shall mean a website, online service, online application, or mobile application, that offers or provides users with addictive feeds as a significant part of the provision of such website, online service, online application, or mobile application.

Compliance Obligations 3 obligations · click obligation ID to open requirement page

CP-01 Deceptive & Manipulative AI Conduct · CP-01.2 · Deployer · Social Media

Gen. Bus. Law § 1510

Plain Language

Operators of addictive social media platforms must provide all users with four user-accessible control mechanisms: (1) an option to turn off algorithmic recommendations entirely; (2) an option to turn off notifications related to the addictive feed, at minimum allowing users to disable notifications altogether or specifically during midnight to 6 AM Eastern; (3) an option to turn off autoplay of embedded media; and (4) a tool allowing users to set a hard daily time limit on platform access — a mere time-spent reminder does not comply. All four mechanisms must be offered as a condition of operating the platform in New York. The definition of 'algorithmic recommendation' contains extensive carve-outs for subscription-based feeds, search results, direct messages, sequential content, and accessibility-related prioritization, which narrow the scope of what must be toggleable.

Statutory Text

It shall be unlawful for an operator to provide an addictive social media platform to a user in this state unless such platform offers mechanisms through which a user may: 1. Turn off algorithmic recommendations; 2. Turn off notifications concerning an addictive feed, provided further that such mechanism shall, at a minimum, provide the user with the ability to turn off notifications overall or to turn off notifications between the hours of 12 AM Eastern and 6 AM Eastern; 3. Turn off autoplay on such platform; and 4. Limit such user's access to such platform to any length of day specified by such user, provided further that any mechanism which solely reminds such user of time spent on a platform rather than allowing such user to limit such user's access shall not be in compliance with this subdivision.

CP-01 Deceptive & Manipulative AI Conduct · CP-01.3 · Deployer · Social Media

Gen. Bus. Law § 1511(1)

Plain Language

The four user control settings required under § 1510 must be presented clearly and accessibly on the platform. Operators are prohibited from deploying any mechanism or design that intentionally subverts user choice, inhibits the purpose of the statute, or makes it harder for users to exercise their rights under the required settings. This is a broad anti-dark-pattern prohibition covering deceptive defaults, hidden toggles, confusing UI flows, manufactured friction, and any other design choice that undermines the user controls the statute requires. The 'intentionally' modifier means the prohibition targets purposeful design decisions, not accidental UX friction.

Statutory Text

The settings required in section fifteen hundred ten of this article shall be presented in a clear and accessible manner on an addictive social media platform. It shall be unlawful for such platform to deploy any mechanism or design which intentionally inhibits the purpose of this article, subverts user choice or autonomy, or makes it more difficult for a user to exercise their rights under any of the prescribed settings in section fifteen hundred ten of this article.

CP-01 Deceptive & Manipulative AI Conduct · CP-01.3 · Deployer · Social Media

Gen. Bus. Law § 1511(2)

Plain Language

Operators may not use dark patterns or any intentional design mechanism that makes it more difficult for a user to deactivate, reactivate, suspend, or cancel their account or profile. This is a standalone prohibition applying to account lifecycle management — separate from the anti-dark-pattern rule protecting the § 1510 settings. It covers both making it harder to leave the platform (deactivate/cancel) and making it harder to return after a break (reactivate), ensuring friction-free account management in both directions.

Statutory Text

It shall be unlawful for an addictive social media platform to deploy any mechanism or design which intentionally serves to make it more difficult for a user to deactivate, reactivate, suspend, or cancel such user's account or profile.