Key takeaways
- Connecticut Governor Ned Lamont signed Senate Bill 5 into law in May 2026, creating an omnibus regulatory framework for artificial intelligence.
- The legislation imposes specific obligations on entities that sell or deploy AI technology within the state.
- Regulated areas include consumer disclosures, employment tools, and broad developer obligations.
- The law applies across a wide range of use cases rather than targeting a single category of artificial intelligence.
The Enactment
Connecticut enacted the measure as Public Act No. 26-15, signed by Governor Ned Lamont on May 27, 2026.
In May 2026, Connecticut Governor Ned Lamont signed Senate Bill 5 into law, establishing a regulatory framework for artificial intelligence usage in the state. Formally referred to as an omnibus law concerning online safety and AI responsibility, the legislation imposes requirements on entities that deploy or sell artificial intelligence technology. The measure covers various aspects of artificial intelligence, including consumer disclosures, employment tools, and developer obligations. Notably, the legislation applies to a wide range of use cases beyond a single category of AI, marking a significant expansion of state-level oversight over automated systems.
Why It Matters
State-level action on artificial intelligence creates immediate compliance obligations for technology vendors and employers operating within the jurisdiction. Because the legislation applies to a wide range of use cases beyond a single category of AI, companies cannot rely on narrow industry exemptions to avoid scrutiny. The inclusion of employment tools means human resources departments must evaluate the software they use for hiring, firing, and workforce management.
This broad approach signals a shift toward comprehensive state regulation, mirroring the early development of state-level data privacy laws. By targeting both the entities that sell the technology and those that deploy it, the statute ensures that liability and compliance duties attach at multiple points in the commercial supply chain. Businesses that purchase off-the-shelf software will need to demand compliance guarantees from their vendors, while developers must ensure their products meet the state's baseline safety and transparency standards before entering the market.
Who Should Care
For lawyers
Counsel advising technology developers, vendors, and employers must immediately audit their clients' software inventories. Employment attorneys specifically need to review how clients utilize automated tools for screening applicants or evaluating current workers. The law requires advising on both developer obligations and consumer disclosures, meaning compliance strategies must span multiple practice areas, including labor, consumer protection, and technology transactions. Attorneys negotiating software-as-a-service agreements will need to draft new indemnification clauses to address the specific statutory duties placed on sellers versus deployers.
For consumers and employees
Workers and the general public gain visibility into when and how artificial intelligence affects their daily lives and economic opportunities. The law mandates consumer disclosures, ensuring individuals know when they interact with automated systems rather than human beings. Employees subjected to algorithmic evaluations will see changes in how companies deploy these tools, potentially gaining more transparency regarding how hiring or promotion decisions are made.
Legal Background
Before this enactment, artificial intelligence operated largely without targeted statutory constraints at the state level. Legislatures previously focused on data privacy or specific automated decision-making in narrow contexts, leaving a regulatory vacuum regarding the broader deployment of generative and analytical algorithms.
While other states have debated similar measures, Connecticut chose to enact an omnibus framework. This approach consolidates rules for online safety and AI responsibility into a single statutory scheme. Historically, attempts to regulate technology have struggled to keep pace with software development, often resulting in fragmented rules that apply only to specific sectors like healthcare or finance. The move toward an omnibus model represents a legislative strategy to capture the technology's general capabilities rather than its specific applications.
What the Legislature Did
The Connecticut legislature drafted Senate Bill 5 to cover various aspects of artificial intelligence, specifically targeting consumer disclosures, employment tools, and developer obligations. Rather than limiting the scope to high-risk systems or specific industries, lawmakers ensured the legislation applies to a wide range of use cases.
The statute places affirmative duties on entities that sell artificial intelligence technology, requiring them to meet baseline developer obligations before their products reach the market. Simultaneously, the law regulates the businesses that deploy these systems. By addressing both online safety and AI responsibility, the legislature created a dual-pronged approach. The inclusion of consumer disclosures ensures transparency at the point of interaction, while the regulation of employment tools addresses the hidden algorithms that dictate economic outcomes for workers.
How It May Be Applied
Regulators will likely scrutinize how companies implement the required consumer disclosures, focusing on whether the notices are sufficiently clear to the average user. Vendors that sell artificial intelligence tools may need to rewrite their terms of service to allocate liability between developers and deployers, likely leading to heavy negotiations over indemnification.
Employers face immediate questions about which specific software qualifies as an employment tool under the statute. For example, standard resume-filtering software may now trigger compliance obligations that previously applied only to advanced predictive analytics. Courts and state agencies will need to interpret the boundaries of these definitions as enforcement begins. Furthermore, the broad application across a wide range of use cases means that non-technology companies—such as retailers, manufacturers, and hospitality businesses—will find themselves subject to AI regulations simply by purchasing modern enterprise software.
Regulatory Comparison
| Regulatory Area | Pre-Enactment Status | Senate Bill 5 Requirement |
|---|---|---|
| Consumer Disclosures | General consumer protection laws applied without specific tech mandates | Specific disclosures required for artificial intelligence interactions |
| Employment Tools | Limited statutory oversight of algorithmic hiring | Regulated deployment of AI in workforce management |
| Developer Obligations | Contract-driven liability between vendor and buyer | Affirmative statutory obligations for developers and sellers |
| Scope of Regulation | Fragmented across specific industries or nonexistent | Omnibus framework covering a wide range of use cases |
The Bottom Line
Plain-English Callout: Connecticut's new law means that if a company makes or uses artificial intelligence, it has to follow strict new rules. Whether a business is selling an AI program or a local employer is using software to sort resumes, the state now requires transparency and accountability. By passing this omnibus law, the state guarantees that consumers will know when they are dealing with automated systems and ensures that developers cannot pass all the blame to the businesses that buy their products.
This article is general legal information and commentary about legal developments. It is not legal advice, does not address your specific situation, and is not a substitute for advice from a licensed attorney. Reading this article and contacting us through this website do not create an attorney-client relationship.
Sources & authorities
- Conn. Pub. Act No. 26-15, "An Act Concerning Online Safety" (Substitute S.B. No. 5), signed May 27, 2026 — source
Further reading
Additional perspectives (a link is not an endorsement):