News March 31 2026
March 31, 2026 — Daily digest of AI law developments.
This article consolidates 2 news stories from March 31, 2026.
Contents
1. Oregon SB 1546 AI Companion Law 2. Reddit v Anthropic Remand
Oregon SB 1546 AI Companion Law
Oregon SB 1546 (the AI Companion Safety Act)[1] is an Oregon state law signed by Governor Tina Kotek on March 31, 2026, that establishes comprehensive regulation of AI companion systems.[2][3] The law takes effect on January 1, 2027.[3]
Overview
Oregon SB 1546 is the first state law regulating AI "companions" — AI systems designed to simulate sustained human-like platonic, intimate, romantic, or companionship relationships by retaining contextual information and personalizing interactions.[3][4] It passed both chambers on March 5, 2026, with near-unanimous support (28-2 in the Senate) and no public opposition testimony.[3][5]
Key Provisions
Disclosure Requirements
Operators must clearly and conspicuously disclose that the service is AI-powered if a reasonable person might believe they are interacting with a human.[4][1]
Safety Requirements
All AI companion operators must:
- Detect suicidal ideation or self-harm in user communications
- Interrupt conversations when risk is detected
- Provide crisis referrals (e.g., suicide prevention hotlines)[3][5]
Minor Protections
When operators have reason to believe a user is under 18:
- Provide hourly reminders to take breaks and that the companion is AI (no misrepresenting as human)
- Avoid sexually explicit content or techniques that foster emotional dependency[3][2]
Transparency and Reporting
Operators must file annual reports publicly online in Oregon, disclosing usage data and safety metrics.[3][4]
Enforcement
Notably, SB 1546 includes a private right of action with $1,000 statutory damages per violation, making it the first chatbot regulation law to allow individual lawsuits.[3][5]
Relationship to Other Laws
Oregon SB 1546 builds on California's SB 243 but adds mandatory crisis response interruptions and minor-specific protections, establishing a consumer protection and public health framework distinct from other state chatbot laws.[3][4]
Significance
Oregon SB 1546 is significant because it:
- Is the first AI companion law to include a private right of action
- Establishes mandatory crisis detection and intervention requirements
- Creates specific protections for minors interacting with AI companions
- Requires public transparency reporting from AI companion operators
- Takes effect January 1, 2027
References
- ↑ 1.0 1.1 Oregon Legislative Information, SB 1546 Overview
- ↑ 2.0 2.1 Transparency Coalition, Oregon Lawmakers Pass Major Chatbot Bill
- ↑ 3.0 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 Baker Botts, Oregon SB 1546: The First Chatbot Law With Real Teeth
- ↑ 4.0 4.1 4.2 4.3 Miller Nash, Oregon's New AI Companion Law: What You Need To Know
- ↑ 5.0 5.1 5.2 Troutman Pepper, Oregon Legislature Passes Bill Regulating Consumer-Facing Interactive AI With Private Right of Action
See individual article: Oregon SB 1546 AI Companion Law
Reddit v Anthropic Remand
Reddit, Inc. v. Anthropic PBC — On March 31, 2026, the U.S. District Court for the Northern District of California granted Reddit's motion to remand its lawsuit against Anthropic to California state court, rejecting Anthropic's argument that federal copyright law preempted Reddit's state-law claims.[1][2]
Reddit sued Anthropic alleging unauthorized scraping of its platform content since approximately 2022 to train Claude and other AI models, in violation of Reddit's user agreements and California law. Anthropic removed the case to federal court, arguing that Reddit's claims were preempted by the federal Copyright Act under 17 U.S.C. § 301.[3]
The court held that each of Reddit's claims — breach of contract, unjust enrichment, trespass to chattels, tortious interference with contract, and violations of the California Business and Professions Code — contained "extra elements" qualitatively different from copyright's exclusive rights. These extra elements include breaching user agreement restrictions on access methods, bypassing technical safeguards, and impairing infrastructure. Under the two-part § 301 preemption test, while Reddit's content falls within copyright's subject matter, the extra elements mean the rights are not equivalent.[1][2]
The remand means the case proceeds in California state court, where Reddit can pursue its claims without the overlay of federal copyright law. Anthropic may appeal the remand order, and may still raise fair use defenses in state court.[1]
Significance
This ruling is significant for AI litigation because it establishes that AI training data cases based on state-law theories (contract, tort, unfair competition) are not automatically preempted by federal copyright law. This allows plaintiffs to keep cases in state court and pursue broader remedies than copyright alone provides. It also means AI companies cannot automatically federalize scraping disputes by invoking copyright preemption.[1]
See Also
References
See individual article: Reddit v Anthropic Remand