Civil rights advocates and lawmakers are urging Connecticut to reform its Freedom of Information laws to ensure transparency regarding the state’s use of algorithms and artificial intelligence. A recent report highlights significant concerns that automated decision-making systems in education and hiring may perpetuate long-standing inequities due to flawed data or inherent bias. The primary obstacle to accountability is the current lack of clear information, as agencies often deny public records requests by citing trade secret exemptions, leaving citizens unaware of how critical decisions affecting their lives are made. This issue is highly relevant to open data because it underscores the necessity of algorithmic transparency for democratic oversight. Without access to the underlying logic and datasets of government software, the public cannot verify whether automated processes are fair or discriminatory. True open data initiatives must extend beyond raw numbers to include the explanatory context of automated systems, ensuring that the public understands the criteria used in public administration rather than relying on opaque "black box" technologies. To address these gaps, a bipartisan bill is advancing that mandates a public inventory of automated systems and establishes guidelines for their ethical use. While this legislative move aims to create necessary guardrails before reliance on AI becomes entrenched, it does not currently resolve the specific FOI exemptions hindering access to technical details. Proponents argue that establishing a comprehensive understanding of how these tools function is essential for building trust and ensuring that data-driven governance serves the public interest equitably.
Source: ctnewsjunkie.comPublished on 2023-04-26