Colorado has taken the lead in implementing regulations for artificial intelligence (AI) systems that have an impact on hiring, housing, and medical decisions. A total of seven bills have been proposed, with one already passed and approved, with the aim of addressing the biased approach of AI in making important decisions.
Despite facing opposition, Governor Jared Polis reluctantly signed the bill on Friday. This bill mandates organizations to evaluate the risk of discrimination caused by AI, establish monitoring programs, and inform consumers if AI is involved in decision-making.
Similar to bills in Washington and Connecticut, Colorado’s legislation has not been well-received by civil rights groups, the tech industry, cautious lawmakers, and concerned governors. These bills aim to tackle AI bias in critical areas such as employment, housing, and healthcare.
There has been significant resistance to the call for regulation due to the complexity of the technology and the potential limitation on its development. Governor Polis also expressed concerns about the bill, emphasizing that it should not hinder AI development.
Colorado’s legislation, along with the other six bills, requires organizations to assess the potential bias of AI systems. Companies that use AI in their operations are required to notify clients when significant decisions involving them are made by AI. This approach aims to increase transparency and accountability in AI applications.
This initiative comes at a time when there are extensive legislative discussions, with over 400 AI-related bills introduced this year. These bills primarily focus on specific aspects of AI, such as deepfake and AI pornography. However, these seven ambitious bills concentrate on discrimination, which is one of the most challenging and universal concerns related to AI.
Rumman Chowdhury, an artificial intelligence specialist for the U.S. Department of State, highlighted that current legislation against discrimination does not adequately address the issues arising from the use of AI. Civil rights attorney Christine Webber pointed out that AI has the potential to amplify prejudice on a much larger scale than a single decision-maker.
Webber’s case provides a clear example of how AI can perpetuate discrimination, such as an AI tool for rental applications giving lower scores to black and Hispanic individuals. This issue also extends to the medical field, where AI systems have been observed to deny specialized care to Black patients based on biased evaluations. Colorado’s bill seeks to address these concerns by implementing annual assessments of AI systems for bias, internal regulation, and reporting of discriminatory results to the state attorney general.
While the intention of the bill is to promote transparency and accountability, there are concerns within the AI industry. While larger AI companies generally approve of these measures, smaller companies see them as potentially too complex. Logan Cerkovnik, founder of Thumper.ai, stated that AI companies also succeeded in adding a clause to prevent anyone other than the state attorney general from suing over the law.
The action taken by the state of Colorado comes as Congress considers federal AI rules. Although Congress has yet to pass any bills regulating AI, Senate Majority Leader Chuck Schumer recently released a report with guidelines for AI regulation.
Source: Yahoo News UK
