HB 1385 requires health insurers and pharmacy benefit managers using artificial intelligence (AI) for medical treatment reviews (utilization review) to include licensed healthcare professionals in key evaluations. Specifically, it mandates that AI decisions must be reviewed by a licensed professional who can question, modify, or override AI determinations based on a patient’s full medical history and individual circumstances. The bill also requires quarterly reviews of AI performance using human assessments of real-world health outcomes to improve accuracy and safety. These provisions apply to all AI tools used in utilization review processes within Maryland’s health insurance system, effective October 1, 2026.
SB 989 (Safe Kids Act) creates a civil cause of action allowing parents or legal guardians of minors depicted in unlawful child exploitation material to sue online platforms that knowingly distribute such material. The bill defines "unlawful child exploitation material" to include AI-generated images that appear indistinguishable from real children, criminalizing platforms that knowingly issue, sell, or distribute these materials. It establishes remedies including actual damages, attorney fees, and punitive damages for affected families, while exempting platforms acting solely as transmission services. The law applies to social media, apps, and websites that create or distribute content featuring minors in sexual situations, including AI-generated depictions.
HB 1399 requires credit bureaus and other consumer reporting agencies in Maryland to meet strict standards when using algorithms to evaluate credit data. It mandates that agencies explain algorithmic decisions in plain language (meeting an 8.0 Flesch-Kincaid readability score), maintain public algorithm registries, and achieve error rates below 0.5% and discriminatory bias below 0.1% when compared to human reviews. Agencies must also conduct quarterly bias audits by independent third parties, implement data governance with minimum data point requirements (1,000-10,000 per category), and require human review within 24 hours before final credit decisions. This bill directly affects agencies that compile credit reports for lenders, insurers, or employers, aiming to ensure transparency and fairness in automated credit evaluations.
HB 1475 requires merchants selling goods or services to disclose when prices are set using algorithms that analyze personal data, mandating a clear statement: "THIS PRICE WAS SET BY AN ALGORITHM USING YOUR PERSONAL DATA." It also prohibits residential landlords or property managers from using software, data analytics, or algorithms to coordinate rent adjustments or lease terms across multiple properties. Violations are treated as unfair trade practices under Maryland’s Consumer Protection Act, subject to enforcement and penalties. The bill directly affects businesses using dynamic pricing and rental property owners/managers utilizing algorithmic tools for pricing decisions.
HB 795 requires Maryland health insurance companies to provide human review for any coverage denial made using artificial intelligence, algorithms, or software tools. Insurers must also report quarterly aggregated data on these AI-related denials, including claim types, member demographics (race, gender, profession), and policy details (individual, group, or exchange plans). These reports must be submitted to the state insurance commissioner and include the total number of AI-related grievances reviewed. The bill directly affects health insurers operating in Maryland and impacts members who face AI-driven coverage decisions.
SB 720, the "Artificial Intelligence Ready Schools Act," requires Maryland’s State Department of Education to create and maintain online guidance for schools, educators, parents, and students on the safe, ethical, and equitable use of AI in K-12 education. It mandates county school boards to adopt AI policies within 120 days of guidance release, designate AI coordinators, and procure AI tools aligned with state standards. The bill also requires Morgan State University to annually certify AI tools and establishes statewide teacher training in AI literacy by July 2027, with compensation for educators participating. Local school systems must integrate AI into workforce preparation standards by June 2027, coordinated with the Maryland Center for Computing Education. The law directly affects all Maryland public schools, educators, students, and technology vendors supplying educational AI tools.
HB 883 prohibits AI developers from making or causing AI to make claims that the AI is a behavioral health provider or can deliver behavioral health care services. It requires AI sold to Maryland consumers to include clear notices stating users are interacting with AI (not a human) and to detect suicidal thoughts or self-harm, automatically referring users to crisis services. Violations carry civil penalties up to $1 million per offense, with funds directed to Maryland’s Behavioral Health Workgroup Investment Fund. The law directly affects AI developers and sellers operating in Maryland, focusing on preventing misleading AI interactions in mental health contexts.
SB 387 prohibits large food retailers (15,000+ sq ft) in Maryland from using dynamic pricing (real-time price changes based on demand or AI) or consumer surveillance data (like location or biometric tracking) to set prices for individual customers. It also bans using protected class data (e.g., race, gender) to deny accommodations or advantages to consumers. The bill further prevents retailers from reducing union-guaranteed employee benefits without negotiating with union representatives. These provisions aim to regulate pricing practices, prevent discriminatory data use, and protect collective bargaining agreements, with violations subject to enforcement under Maryland’s consumer protection laws.
HB 9 establishes a Maryland 3-1-1 Oversight Board to manage a statewide program expanding non-emergency service access through AI-powered tools. The bill requires all Maryland counties to implement AI chatbots (by June 2027) and voicebots (by December 2028) that provide multilingual support, route calls to appropriate services using geographic data, and escalate complex requests to live agents. These systems must align with accessibility and equity standards, with counties without existing 3-1-1 systems required to adopt the program by July 2028. The Oversight Board, composed of state officials and county representatives, will oversee vendor selection, performance monitoring, and statewide data standards.
HB 712 establishes new rules for holding AI developers and deployers liable when their systems cause harm. It creates a legal path for people harmed by "high-impact" AI systems - like those used in criminal justice, housing, or medical devices - to sue for defective design, lack of warnings, or broken promises. The bill includes key protections: if developers followed safety testing, courts would assume the AI was safe unless proven otherwise, and small businesses (under 20 employees or 10,000 users) are exempt. It also removes "contributory negligence" as a defense in such cases and allows the state attorney general to sue for widespread harm.