(Bloomberg) -- The US needs to shield Americans from the risks posed by artificial intelligence while promoting the emerging technology with at least $32 billion in annual government spending to stay ahead of rivals like China, according to a highly anticipated policy blueprint from a bipartisan group of senators.

Congress should craft legislation that mitigates AI’s potential harm and funds research to boost US economic and national security, according to the plan released Wednesday by Senate Majority Leader Chuck Schumer. US companies including OpenAI, Alphabet Inc.’s Google, and Meta Platforms Inc. currently hold the global lead, yet China’s race to develop highly advanced systems is putting pressure on lawmakers to craft a more lasting approach.

Funding will “keep our companies, our universities, our workers at the cutting-edge and cement America’s dominance in AI, including out-competing the Chinese government, which we know is putting lots of dollars into this area already,” Schumer said in unveiling the 30-page document.

The blueprint culminates more than a year’s worth of activity on Capitol Hill to familiarize senators with AI, a first step toward eventually writing legislation governing the rapidly evolving technology. The senators last year held a series of closed-door forums featuring labor and tech industry leaders, including OpenAI’s Sam Altman, Google’s Sundar Pichai, Meta’s Mark Zuckerberg and Elon Musk of Tesla Inc., to examine AI’s vast implications for everything from national security to jobs to individual privacy.

Rapid developments in AI have fueled concerns that the technology could foment the spread of misinformation and expose consumers’ personal data. Executives from Google, Microsoft, IBM and OpenAI have urged more federal oversight to help guarantee safety, and three dozen venture capital firms agreed to abide by voluntary commitments after President Joe Biden in October signed an executive order that set initial safeguards and built on earlier undertakings from industry.

Policymakers in Washington have been seeking to build on that effort, seizing on the promise of AI to advance scientific discovery and solve global crises while also working to limit its dangers. Wednesday’s framework is designed to strike that regulatory balance and jumpstart legislative action, the senators said.

Their recommendations, especially on funding, will be a heavy lift to get through a bitterly divided Congress. House Speaker Mike Johnson faces unrelenting pressure from conservative Republicans to cut spending, and lawmakers historically have a poor track record of regulating new technologies, further dimming prospects for results — especially during an election year.

Yet framing their plan as a way to counter Beijing may help their cause. Challenging China over economic and national security has been a rare area of common ground for lawmakers, most recently in a new law that passed with wide bipartisan support requiring TikTok’s Chinese owner ByteDance Ltd. to sell the popular video-sharing app or face a ban in the US.

Schumer struck an optimistic tone as he briefed reporters ahead of the plan’s release, saying he intends to meet with Johnson in coming weeks to find a path forward. His co-authors — Republicans Mike Rounds and Todd Young, along with Democrat Martin Heinrich — said they hoped their proposal sets up action as soon as this year.

Some of the AI blueprint’s goals, such as safeguarding elections from deepfakes and misinformation, could find widespread favor in Congress given that lawmakers from both parties are vulnerable to such manipulation by foreign adversaries. The plan also calls for exploring ways to minimize the impact of job displacement as a result of artificial intelligence, including by retraining workers, while also doing more to shield consumer privacy and minimize bias.

The $32 billion figure for non-defense areas like health and energy was originally recommended in 2021 by the National Security Commission on AI, an independent board led by former Google CEO Eric Schmidt. A chunk of that funding would go toward agency-led research programs to foster more innovation. Promoting the use of AI in health-care is another area the senators urged fellow lawmakers to further explore.

US efforts to create guardrails lag behind other governments including the European Union, where the parliament in March approved the most comprehensive regulation of AI in the Western world. The EU measure bars using AI to detect emotions in workplaces and schools, while limiting its deployment in high-stakes situations like sorting job applications.

In addition to the Schumer-led blueprint, lawmakers have released several other bipartisan proposals that would, for example, bolster the government’s use of AI or protect artists from being mimicked by AI without their consent. The senators’ roadmap directs congressional committees to finetune and advance AI legislation, and calls out a handful of existing bills they support, such as one that would create a shared national research hub for educators and students to develop the tech.

(Adds link to document and box summarizing highlights)

©2024 Bloomberg L.P.