Incentivizing AI research
A bipartisan House measure introduced on Wednesday looks to create federally-funded research competitions to enhance AI security and understanding of the technologies.
Led by Reps. August Pfluger, R-Texas, and Josh Harder, D-Calif., the Reliable Artificial Intelligence Research Act would create two national competitions operated by the Department of Homeland Security focused on AI interpretability and adversarial robustness. The legislation would set aside $10 million in prizes to award through the national competitions over the next five years.
“As AI becomes more deeply woven into our economy and our national defense, we cannot afford to leave it vulnerable to those who would exploit it,” Pfluger said in a statement. “Adversaries are already looking for ways to manipulate these systems, and we must stay ahead of them. The Reliable Artificial Intelligence Research Act puts American ingenuity to work on making AI more transparent, more predictable, and more resistant to attacks.”
Sens. Jim Banks, R-Ind., and Maggie Hassan, D-N.H., previously introduced a companion bill in the upper chamber.
Stopping federal use of biometric tech
A group of House and Senate Democrats reintroduced legislation last Friday that would prohibit the federal government from using biometric technologies. The lawmakers said their proposal is in response to Immigration and Customs Enforcement and Customs and Border Protection’s use of a mobile facial recognition app, which they said has been used to target immigrants.
The measure is sponsored by Sens. Ed Markey, D-Mass., and Jeff Merkley, D-Ore., and Reps. Pramila Jayapal, D-Wash., Rashida Tlaib, D-Mich., and Ayanna Pressley, D-Mass.
The bill would, in part, “place a prohibition on the use of facial recognition technology by federal entities, which can only be lifted with an act of Congress” and “prohibit the use of federal dollars for biometric surveillance systems.”
“Every American who values their right to privacy, stands against discrimination, and believes people are innocent until proven guilty should be concerned about facial recognition technology fueling a national surveillance state,” Merkley said in a statement. “Enacting a federal moratorium on this technology is critical to ensuring we can rein in inappropriate surveillance and protect our communities’ fundamental right to privacy.”
Safeguarding data collected by AI-powered cameras
Sen. Josh Hawley, R-Mo., announced on Wednesday that he is introducing legislation to create guardrails around the use of AI-powered cameras to protect Americans’ data that is collected by the technologies.
The senator’s measure comes after the Senate Judiciary Subcommittee on Crime and Counterterrorism — chaired by Hawley — held a hearing last month that examined the use of automatic license plate readers from companies like Flock Safety.
Hawley’s bill, the Stop Flock Abuse Act, would prohibit local governments from selling or sharing collected data with non-governmental third parties, prevent users from abusing access to the systems, preserve an audit log of network access and create data security and retention guidelines.
“AI-powered cameras produced by Flock and its competitors are tracking and creating digital footprints of everyday Americans — a nightmare for individual liberties unless Congress enacts meaningful rules and accountability,” Hawley said in a statement.
Holding AI firms liable for hacking incidents
Hawley and Sen. Chris Murphy, D-Conn., also announced on Thursday that they are introducing a proposal that would hold AI agent operators and developers criminally and civilly liable for instances in which their advanced models hack into other systems or networks.
The announcement came after Hawley chaired a Senate Homeland Security and Governmental Affairs Subcommittee on Disaster Management, District of Columbia and Census hearing on Wednesday that looked into instances where AI tools broke out of their testing environments and attempted to hack into other companies or networks, including websites hosted by federal agencies.
“Hacking is a crime, and when AI agents conduct dangerous cyberattacks, the corporations and executives responsible for those AI agents need to be held accountable,” Murphy said in a statement. “Our bipartisan bill forces the heads of big AI companies to develop responsibly or face prison time for the damage done by their products to everyone else.”
Prohibiting human-like chatbots from interacting with minors
House lawmakers introduced a measure on Thursday that would prohibit AI chatbots from communicating by default in a human-like manner with minors, including through the use of simulated emotions or relationship-building responses.
The bipartisan bill is sponsored by Reps. George Whitesides, D-Calif., Mike Kennedy, R-Utah, Doris Matsui, D-Calif., and Mariannette Miller-Meeks, R-Iowa.
The legislation would, in part, require that chatbots engaging with children refrain from claiming to be human or pose as “credentialed professionals like doctors or counselors,” periodically state that they are not real and do not “guilt-trip” children into continuing to chat with them. The proposal also calls for the National Institute of Standards and Technology to work with industry and academic partners “to develop testing standards and examples of compliant behavior, and companies that follow the guidance won’t be penalized.”
The lawmakers noted that their measure would not affect products or features for adults and would also not require age verification.
“Parents should not have to worry that an AI chatbot is pretending to be their child’s friend or encouraging them to form an emotional attachment to a machine,” Miller-Meeks said in a statement. “The Protecting Kids from Human-Like Chatbots Act puts parents in control and makes sure AI chatbots used by minors are designed to function as tools, without forming fake relationships. This bill is about giving families peace of mind and making sure new technology serves our children without trying to replace real human relationships. Protecting our kids is a priority we should all share.”
Keeping data centers off federally-owned public lands
Sen. Ron Wyden, D-Ore., introduced legislation on Wednesday that would ban the construction of data centers on federally-owned public lands. The measure would also cover the sale or transfer of relevant public lands for the development of new data centers and related infrastructure.
“Public lands belong to the taxpayers, not corporations, and especially not [President] Donald Trump,” Wyden said in a statement. “If there’s one thing I’ve learned during my time in public service, it’s that when you come for Oregonians’ public lands, you’ll have hell to pay. My legislation is the ironclad path to ensuring that Trump – or any future administration – cannot hijack our treasured public lands to build data centers that use up water resources and jack up energy costs.”
More transparency in data center use and development
Sens. John Curtis, R-Utah, and Lisa Blunt Rochester, D-Del., also introduced two measures focused on enhancing transparency around the construction of new data centers.
One proposal, the Data Center Community Empowerment Act, would create a program within the Energy Department to provide technical assistance to state, local and Tribal governments on the potential impacts of data center development in their regions, create a “Data Center Community Toolkit” and provide independent project analyses for interested communities.
The second bill, the Data Center Transparency Act, would similarly require that the Environmental Protection Agency and Energy Information Administration collect and update information related to the environmental impacts and resource uses of data centers.
“Americans are rightly skeptical when a large company tells them not to worry about the impacts of a new data center in their community,” Curtis said in a statement. “The answer to that distrust is transparency. Communities deserve independent, reliable information about what these projects mean for their power bills, water, infrastructure, and local economy. These bills give local leaders the facts and expertise they need to ask tough questions and make informed decisions — without Washington or industry making those decisions for them.”
Establishing an AI Safety Board
Sens. Mark Warner, D-Va., Brian Schatz, D-Hawaii, and Andy Kim, D-N.J., introduced legislation on Tuesday that would create an Artificial Intelligence Safety Board within the Commerce Department to convene representatives from agencies across government — including from the Cybersecurity and Infrastructure Security Agency, the National Security Agency and the Treasury Department — to evaluate emerging AI risks. This board would also develop enforceable safety and security standards for the most advanced AI tools.
The measure, in part, would also create a national AI incident database, require AI companies to report serious safety and security incidents within certain timeframes and create “ secure federal testing environments that can use resources at the National Security Agency and Department of Energy National Laboratories to conduct sensitive pre-deployment testing of frontier models.”
In a statement ahead of the bill’s formal introduction, Schatz said the proposal “is about making sure humans remain in control of AI models and preventing future breaches with rigorous standards, testing, and oversight. Immediate risk requires immediate action.”
Creating new federal oversight of data collection
Sen. Kirsten Gillibrand, D-N.Y., also reintroduced a bill on Wednesday that would stand up a new federal agency to help protect Americans’ data.
Gillibrand’s measure, the Data Protection Act, would also limit the collection, sharing and use of personal data; provide oversight of high-risk data practices and data collection for commercial purposes; maintain a public list of data aggregators; and examine “the social, ethical, economic, and civil rights impacts of data collection practices.”
In a statement, Gillibrand said her measure would “create an independent federal agency to protect data and privacy, equipped with the resources to give Americans the meaningful protection they deserve.”
Creating a centralized website of public HHS data
Rep. Cliff Bentz, R-Ore., also introduced legislation on Monday that would direct the Department of Health and Human Services to establish “a centralized repository of all publicly available data.”
The congressman said his measure is designed to provide easier access to HHS’ collection of public information that is currently broken down into “tens of thousands of distinct data sets across multiple websites.”
In addition to directing HHS to create a centralized website where all of the public data can be accessed, HHS would be required to report to Congress on the types of datasets that are available, how frequently they are updated and the barriers that prevent publication of the information.
“We all want to improve our healthcare system, and access to data is absolutely essential to these efforts,” Bentz said in a statement. “My Public Access to HHS Health Data Act will help all groups more quickly access this essential information.”
