By Armando DeChalus
Artificial intelligence (AI) continues to be a central focus for investment advisers, with much of the discussion centered on vendor oversight, cybersecurity, privacy, and recordkeeping. However, a new wave of state regulation is highlighting another area for RIAs to watch closely : the use of AI and automated tools in employment-related decisions.
Connecticut recently enacted Senate Bill 5, a broad AI law that includes new requirements governing the use of certain automated employment related decision processes.1 The employment provisions are focused primarily on transparency and disclosure. Beginning October 1, 2027, employers deploying covered automated technologies in Connecticut must provide notice when employees or applicants are interacting with such technology, unless that interaction would be obvious to a reasonable person. Employers must also provide written notice before using automated technology as a substantial factor in certain employment related decisions.2
The law is notable because of how broadly the legislation defines “covered technology”.3 Rather than focusing solely on traditional AI hiring tools, Connecticut’s framework can apply to automated processes that generate rankings, scores, recommendations, or other outputs that meaningfully influence employment decisions. The law also requires employers to provide certain disclosures and explanations when automated technology plays a significant role in adverse employment decisions. 4
Although Connecticut’s law is not specific to financial services firms, it may be relevant to RIAs that use third party technology in connection with recruiting, employee management, or workforce oversight. Some firms may not view themselves as using AI for “recruiting” in the narrow sense. However, many RIAs rely on platforms that may incorporate automated scoring, ranking, screening, monitoring, analytics, or recommendation features. Examples may include applicant tracking systems, performance management software, employee monitoring tools, or workforce analytics products. As a result, the practical question may not be whether a firm has intentionally adopted an “AI hiring tool,” but whether existing RIAs are already using automated processes that influence employment-related decisions.
Connecticut’s law also fits within a broader trend of states adopting frameworks that address AI and automated decision-making in employment and other consequential decisions. Colorado and California have adopted similar approaches, while other jurisdictions have implemented more targeted requirements. For multi-state RIAs, this growing patchwork may create additional compliance complexity, particularly when vendors deploy AI-enabled features across offices or jurisdictions.
For RIAs, these developments may be best understood through the familiar lens of governance and vendor oversight. Firms already evaluate third party technology vendors for cybersecurity, privacy, recordkeeping, and operational risk. AI-enabled HR and workforce tools may increasingly become part of that same diligence process.
As state regulation continues to develop, firms may wish to inventory employment related technologies, review vendor disclosures, understand whether automated outputs are used in employment decisions, and consider whether existing AI and vendor management policies adequately address these tools.
The core risk may not come from intentionally adopting AI for hiring, but from failing to recognize where AI-driven decision-making is already embedded in existing systems.