Rhode Island Signed Two AI Laws in One Day — and 6 More Moves This Week
Rhode Island enacted AI chatbot mental-health protections, Inver Grove Heights turned back a $150M lawsuit threat to pass a data-center moratorium, a federal court kept the Mobley v. Workday AI hiring-bias case alive, and RealPage's biggest landlord customer agreed to pay $7 million.
Rhode Island had a busy June 22: Governor McKee signed two AI companion-chatbot safety bills into law on the same day — one banning AI from posing as a licensed therapist, another requiring AI companions to detect suicidal ideation and connect users to crisis services. Those two laws are the headliners of a week packed with AI-policy action across all three branches of government: a data-center council that voted 3-2 despite a developer's threat of $150 million in damages, a $7 million consent decree in the landmark RealPage algorithmic-rent case, and a federal judge who refused to dismiss AI hiring-bias claims against Workday. Here are the eight developments you need to know about.
-
NEW LAW
1. Rhode Island banned AI from posing as a licensed therapist or providing psychotherapy
Rhode Island · Signed June 22, 2026Governor Dan McKee signed H 7349/S 2197 on June 22, 2026, making Rhode Island the first state to prohibit AI from providing mental-health therapy or psychotherapy, or from representing itself as capable of doing so. The law limits AI to supporting roles — information retrieval, scheduling, administrative tasks — and reserves the therapeutic relationship for licensed human professionals. Violations are actionable under Rhode Island consumer-protection law.
-
NEW LAW
2. Rhode Island required AI companion chatbots to detect suicidal ideation and refer users to crisis services
Rhode Island · Signed June 22, 2026Also signed June 22, S 2195/H 7350 requires operators of AI companion chatbots — platforms designed for ongoing emotional connection or parasocial relationships — to implement detection protocols for users expressing suicidal ideation or self-harm intent and to route those users to the 988 Suicide and Crisis Lifeline. Operators who fail to maintain these protocols face civil penalties of $15,000 per day payable to Rhode Island's suicide prevention fund. The law is among the first in the country to impose mental-health crisis-safety obligations on AI companion platforms.
-
VETOED
3. Arizona's governor vetoed three AI bills — including a deepfake criminalization measure
Arizona · Vetoed June 19, 2026Governor Katie Hobbs vetoed three AI bills on June 19, 2026 as part of a batch of 88 bills rejected during the state's legislative budget standoff. The highest-profile AI casualty was HB 2133, which would have criminalized the creation or distribution of AI-generated sexually explicit synthetic depictions without consent — described by Courthouse News as 'Arizona governor vetoes bill criminalizing AI deepfakes.' HB 2311 (automated-decision transparency) and HB 2592 (AI in insurance) were also vetoed in the same batch. All three bills now require reintroduction in a future session.
-
COURT
4. A federal judge kept the Mobley v. Workday AI hiring-bias claims alive
United States (N.D. Cal.) · Order on Motion to Dismiss Third Amended Complaint, June 22, 2026Judge Rita F. Lin ruled on June 22, 2026 on Workday's motion to dismiss the Third Amended Complaint in Mobley v. Workday (No. 3:23-cv-00770). FEHA race and sex discrimination claims survived; the ADA disability claim survived; only the Asian American race-discrimination claim was dismissed. California's FEHA claims now proceed alongside federal Title VII, ADA, and ADEA claims. The case — the first major collective action targeting an AI-screening software vendor as an employer-like actor under federal civil-rights law — remains in discovery and continues to shape AI vendor liability doctrine.
-
MORATORIUM
5. Inver Grove Heights, Minnesota voted 3-2 for a data-center moratorium despite a $150 million lawsuit threat
Inver Grove Heights, MN · Effective June 26, 2026 — expires June 26, 2027The Inver Grove Heights City Council voted 3-2 on June 26, 2026 to approve a one-year interim moratorium on new data center development — after the developer behind a pending proposal at the former Travel Tags property on Carmen Avenue threatened civil damages exceeding $150 million if the council moved forward. The council received legal briefing on that exposure and voted anyway. The moratorium, effective immediately, gives the city one year to study zoning, infrastructure, and land-use impacts. The 3-2 vote despite a nine-figure liability threat is one of the sharpest examples yet of a local government accepting legal risk to maintain control over large-scale AI infrastructure.
-
COURT
6. RealPage's largest landlord customer agreed to pay $7 million to nine state attorneys general
United States · LivCor consent decree entered June 26, 2026LivCor LLC — a Blackstone-owned property manager and the largest single landlord defendant in the DOJ's RealPage algorithmic rent-pricing case — agreed to a consent decree entered June 26, 2026, requiring a $7 million payment distributed among nine co-plaintiff state AGs (MN, NC, CA, CO, CT, IL, MA, OR, TN). The main DOJ-RealPage consent decree, entered earlier in 2026, bars RealPage from feeding competitors' real-time pricing data into its algorithm and places the company under a seven-year compliance monitor. The LivCor settlement is the first landlord-level financial resolution in the case; DOJ suits against Greystar and other property managers continue.
-
COURT
7. NAACP sued xAI over unpermitted gas turbines at its Mississippi data center — the first Clean Air Act suit against an AI company
United States (N.D. Miss.) · Litigation ongoing (filed April 14, 2026; DOJ moved to intervene/dismiss June 16)The NAACP filed suit against xAI Corp. in the Northern District of Mississippi (No. 3:26-cv-00074) on April 14, 2026, alleging that xAI operated 27 methane gas turbines without required Clean Air Act permits at its Colossus 2 facility in Southaven, Mississippi — a majority-Black community. The complaint seeks an injunction to shut down the unpermitted equipment. In June 2026, the DOJ moved to intervene and dismiss the suit on sovereign-immunity grounds. The case remains the first Clean Air Act citizen suit filed against an AI data-center operator, with significant implications for how environmental law applies to AI infrastructure.
-
PROPOSED
8. Hawaii's AI chatbot safety bill hits its veto-intent deadline today — Governor Green has until July 15
Hawaii · Awaiting governor action; final deadline July 15, 2026Hawaii SB 3001 CD1 — requiring AI companion platforms to disclose their AI nature, protect minors from manipulative engagement techniques, and maintain crisis protocols for users expressing suicidal ideation — is enrolled and awaiting Governor Josh Green's signature. June 30 is the deadline for the governor to issue a formal intent-to-veto notice; no such notice has been announced as of this writing. Governor Green has until July 15, 2026 to sign or veto; if he takes no action, the bill becomes law without his signature. Hawaii would join Rhode Island and Georgia as states with enacted AI companion-chatbot safety laws in 2026.
This week's common thread is that AI governance is reaching deeper into daily life — housing markets (the RealPage enforcement chain), hiring pipelines (Mobley v. Workday), mental-health services (the Rhode Island dual signing), and environmental justice (NAACP v. xAI). The $150 million threat in Inver Grove Heights that the council voted through anyway signals that communities are becoming more willing to absorb legal risk to maintain local control over AI infrastructure decisions. Meanwhile, the end of the week brings a deadline moment for Hawaii: whether Governor Green acts or stays silent by July 15 determines whether the state's AI companion-safety bill becomes law. We'll update the dataset as soon as the outcome is known.