Mandatory impact assessments for high-risk AI systems before and during deployment. Here is what Connecticut businesses need to know in 2026.
Connecticut has enacted SB 2 — AI Accountability. Developers and deployers of high-risk AI must conduct impact assessments and disclose usage.
State law does not replace federal law — you must comply with both. These federal rules apply to ai risk assessment nationwide:
Connecticut's SB 2 — AI Accountability often includes size-based exemptions. Businesses with fewer than 25 employees may be exempt, but should verify thresholds. Always review the specific statute for employee count and revenue thresholds.
The key deadline in Connecticut is October 1, 2026. The law is enacted and compliance is required by the deadline above.
Connecticut penalties for AI non-compliance: Up to $25,000 per violation. While enforcement is still developing, companies found non-compliant face civil penalties and potential litigation.
Federal law does not currently preempt state AI laws. Connecticut's SB 2 — AI Accountability applies independently of federal rules. Federal laws like ECOA, FCRA, and HIPAA also apply alongside state law — so you must comply with both.
Best practice: document all AI systems used, conduct an internal audit, implement required disclosures, and keep records for at least 3 years. For high-risk uses like ai risk assessment, consider hiring an independent third-party auditor to validate compliance.
Take the free 4-question risk snapshot. See which laws apply, your risk level, and the first actions to take — no signup required.
Take the Free Assessment →