Connecticut has enacted Senate Bill 5, titled An Act Concerning Online Safety, creating one of the most comprehensive state‑level regulatory frameworks governing artificial intelligence, automated decision‑making in employment, and online platform protections for minors. The law generally takes effect on October 1, 2026 and October 1, 2027, and introduces a wide range of obligations affecting employers, technology providers, and digital platforms, along with new enforcement mechanisms and workforce‑development initiatives.
Expanded Regulation of AI in Employment Decisions
SB 5 makes a notable amendment to Connecticut’s discrimination laws by clarifying that an employer’s use of automated employment‑related decision technology (such as AI‑driven hiring tools, screening systems, or algorithmic evaluations) is not a defense to a discrimination claim. In other words, employers remain fully responsible for discriminatory outcomes even if decisions are influenced or made by AI systems. This goes into effect on October 1, 2026.
The law also permits consideration of anti‑bias testing and mitigation efforts as part of the legal analysis, effectively encouraging employers to audit and validate AI tools used in employment decisions. This provision signals increased scrutiny of algorithmic decision‑making and places a premium on proactive bias testing and documentation.
New Disclosure Requirements for AI‑Assisted Employment Decisions
Employers that use automated decision technologies in hiring or employment‑related decisions must provide clear notice to employees and job applicants that such systems are being used. This reinforces transparency requirements and aligns with broader national trends requiring disclosure when AI materially influences employment outcomes. Such disclosure is required beginning October 1, 2027.
Additionally, beginning October 1, 2026, employers must report to the state Department of Labor if layoffs are related to the use of AI.
Regulation of AI Providers and Frontier Model Developers
The law introduces new requirements for AI developers and providers, including:
- Subscription‑based AI services must provide written disclosures of key terms and obtain affirmative consumer consent before entering into or renewing subscriptions.
- Developers of advanced or “frontier” AI models must implement internal compliance systems, including whistleblower protections and reporting processes for potential risks or misconduct.
These provisions aim to promote accountability in the development and commercialization of advanced AI technologies.
New Requirements for Social Media Platforms and Minor Users
SB 5 also imposes significant new obligations on social media platforms, particularly with respect to users under 18. Covered platforms must implement:
- Age‑verification mechanisms;
- Content warnings for sensitive material;
- Restrictions on certain types of notifications; and
- Parental control tools to help monitor and limit minors’ online activity.
These requirements reflect growing concern over the impact of digital platforms on youth and are likely to require substantial operational changes for covered companies.
Additional AI Governance and Consumer Protection Measures
The law also regulates AI companion systems and generative AI tools, bringing them within the scope of consumer protection oversight. Violations related to AI systems, employment decision technologies, or online safety requirements may be treated as unfair or deceptive trade practices, enforceable by the Connecticut Attorney General.
In addition to injunctive relief, the Attorney General may recover investigation costs, expert witness fees, and attorneys’ fees, increasing the financial risk of noncompliance. Frontier AI developers face additional civil penalties for failing to comply with whistleblower protections.
Workforce Development and Education Initiatives
Beyond regulation, SB 5 includes a strong workforce and education component, directing the creation of programs such as a Connecticut AI Academy and integrating AI‑related instruction into teacher preparation and workforce training programs. These provisions signal a parallel effort to prepare workers and educators for an AI‑driven economy while regulating its risks.
Employer and Business Takeaway
Connecticut employers and businesses will face expanded obligations around the use, disclosure, and governance of AI technologies, particularly in employment contexts. Employers using AI in hiring, promotion, or performance evaluation should expect:
- Increased scrutiny of algorithmic tools under anti‑discrimination laws;
- New requirements to disclose AI use to applicants and employees;
- Pressure to conduct and document bias testing and validation efforts; and
- Potential exposure under consumer protection laws for noncompliance.
For technology providers and platform operators, SB 5 introduces broad new compliance frameworks affecting product design, user disclosures, and content controls. Given the law’s breadth and long lead time, organizations should begin identifying AI use cases, reviewing vendor relationships, and building governance frameworks that address transparency, fairness, and regulatory compliance ahead of the various effective dates.