AI experts informed the House Oversight Committee that AI offers an enormous promise for federal workers, but also raises solemn concerns about privacy and security concerns. (Photo by Jennifer Shutt/States Newsroom)
During the congress continues his investigation The role that AI should play in the governmentMembers of the House Oversight Committee question the use and procurement of AI instruments for the work of the government, as well as the data protection that concerns the technology at unregulated level.
At a hearing on June 5, experts testified that the federal government is currently using AI in some capacities, but several factors, such as the antiquated technology systems and data practices from the government, could prevent federal employees from fully exploiting the technology. You also said that of the government Current procurement practices For AI technology, they snail-paced down its introduction that risks other countries that take the leadership in the global AI race.
“The question before us is not whether the Federal Government should accept AI, but whether we will lead or follow,” Bhavin Shah, founder and CEO of MoveWorks, a AI company that summarizes with private companies and local governments.
Many federal departments Use AI According to the current procurement standards such as the Department of Health and Human Services, this was approved, the KI used for outbursts for medical research and persecution of diseases, and the Department of Veterans Affairs, which uses AI to analyze medical records and data for predictive incidents.
In his testimony, Shah said that the employees of the federal government earn the same quick access to AI-powered instruments that the private sector uses to maintain efficiency and cost savings. He said MoveWorks spent three and a half years and $ 8.5 million to reach Fedramp status.
“This is an unaffected barrier for smaller AI innovators who often develop the most modern solutions,” said Shah.
Ylli Bajraktari, President and CEO of the technically oriented thought factory special competition project, said that the United States is still leading the snail-paced introduction of AI into the state use of our “main disadvantage” against countries like China.
“We are hindered by bureaucratic inertia, IT infrastructure and a lack of AI alphabetization in the workforce,” said Bajraktari. “Overcoming these obstacles is the key to winning the global tech competition.”
The suggestions of Bajractari to the committee included the establishment of a Council for AI in the “space race” in the White House, the boost in expenses for AI research and development from non-defenses to $ 32 billion and the introduction of a targeted AI talent strategy, which promote the literacy of AI alphabetization and first-class talent Would attract abroad. He also proposed to revise the procurement process in order to integrate AI into government systems faster and to strengthen global alliances in AI and cyber security.
However, the excessive simplification and the rapid introduction of AI in the government are a great risk to security and data protection, said Bruce Schneier, an employee and lecturer at Harvard Kennedy School. Schneier warned the congress members that Elon Musk’s latest actions – who I just stepped down as a leader of the Department of Government Efficiency (Doge) this week underline the dangers of the unrestricted AI use.
In February, Musk said the Doge team used AI to make decisions About the jobs of federal workers. According to reports, Doge also had access to sensitive data from various federal departments, such as the payment department system for the finance department, social security and other demographic data. In April, 48 legislators expressed their concerns with musk with non -authorized AI systems on these and other data records.
Musk’s actions violate what the committee intended, Schneier said, which is responsible for the release of AI and protects the interests and rights of the Americans. The Americans’ data was consolidated and fed to non-mobile AI models, which represents the national security risks.
“All of you have to assume that our opponents have copies of all data that Doge Peeltrated and the access to all networks has set out that Doge Security Control has removed,” said Schneier. “And your data can be used against you.”
Musk’s actions and access to the data of the Americans and His reported drug usewere the center of a 40-minute debate before the hearing on June 5, in which the Republicans again blocked the Democrats with 21 to 20 votes to try to testify muscle before the congress.
“He reduced our government, endangered Americans and the public service for his own financial profit,” said Rep. Stephen Lynch, a democrat from Massachuse and the incumbent ranking member of the House Oversight Committee.
Linda Miller, founder and Chief Growth Officer at Tracklight, a platform that conducts fraud recognition for government programs, said in her testimony that Doges actions are an example of why it is complex to impose innovations of the private sector.
“My colleagues have some very clever suggestions for charging the RIP and replacing efforts by Legacyit systems and eliminating procurement barriers, but we have to be realistic how capable the government is to take these quick changes, no matter how urgently they are needed,” she said.
Miller said the best use of AI in government work today should be to automate routine processes and repeating tasks that would free the federal employees for more work. She recommended that the congress to accelerate the introduction of AI “regulatory sandboxes” – can consider – controlled environments in which AI systems can be put on trial under the supervisory authorities before being released on the scale.
“Wholesale changes to LEGACY -IT systems and the Federal Audquisition system will take years, but we can create innovative laboratories in which AI projects in the Proof of Concept Regulation Käden can be operated in carefully controlled environments in order to show the art of the possible,” said Miller.
