Artificial Intelligence
plan of action
Thoroughly research and find ironclad solutions for the issues of bias, discrimination, consumer privacy, and the social/ethical implications of A.I. It’s critical we establish ethical frameworks that ensure A.I. enhances our global strength and is beneficial for society overall.
Do not allow the U.S. military to use fully autonomous weapons under any circumstance.
Support the right of private companies to decide what their tech can and cannot be used for. Do not allow the federal government to penalize or punish them for their decisions, like the Trump/Vance administration has tried to do with Anthropic.
It should be up to local and state governments to decide whether to end tax breaks for data center construction. The federal government should not offer tax breaks exclusively for artificial intelligence beyond Research and Development (R&D) Tax Credits.
Make sure experts are consistently judging A.I.’s output reliability, providing for constant independent audits and evaluation.
To help control noise pollution from data centers, resume the funding for the Environmental Protection Agency’s Office of Noise Abatement and Control, an office that was phased out in 1982 but that remains on the books.
Regulating A.I. is tricky because we must balance the many benefits with a large variety of risks – all without stifling A.I.’s progress. But this can be done! Government can absolutely be a stabilizing force here, providing safety measures to prevent harm to Americans without imposing harsh regulations that would stifle competition and be a drag on innovation.
We need a single, comprehensive regulatory framework. Rules and regulations cannot be fragmented; we need consistency. So far, state legislatures have largely taken the lead in regulating A.I., which is understandable given there have been no federal protections in place regarding data security and privacy regulations. However, because we are at such a critical moment – and because this is so incredibly complex – a patchwork of state laws across the country will not cut it.
We must proceed cautiously. The European Union (EU) passed the first comprehensive set of rules for providers, deployers, and distributors of A.I. in the EU market in March 2024. We believe the law goes too far, and the excessive red-tape, stringent risk assessments, and broad requirements for foundation models will likely impede innovation, productivity and progress, as well as act as huge barriers to entry for new players – something we cannot let happen here.