US senators propose guardrails for government AI purchases and operations
(2024/06/13)
- Reference: 1718283609
- News link: https://www.theregister.co.uk/2024/06/13/prepared_for_ai_bill/
- Source link:
Two US senators have introduced a bipartisan bill that defines guardrails for the acquisition and implementation of AI across the federal government.
Senators Gary Peters (D-MI) and Thom Tillis (R-NC) introduced the Promoting Responsible Evaluation and Procurement to Advance Readiness for Enterprise-wide Deployment (PREPARED) for AI [1]Act yesterday. The aim, [2]according to the pair, is to ensure the government can make use of the benefits of AI "while safeguarding against potential risks and harms."
"Artificial intelligence has the power to reshape how the federal government provides services to the American people for the better, but if left unchecked, it can pose serious risks," warned Peters. "These guardrails will help guide federal agencies' responsible adoption and use of AI tools, and ensure that systems paid for by taxpayers are being used safely and securely."
[3]
The [4]bill [PDF] proposes requiring government agencies to conduct a risk assessment exercise before acquiring and implementing AI. Risk audits would be required regularly after systems are implemented.
[5]
[6]
The bill also requires agencies to consider their AI implementation plans with regards to rights and safety of the public and requires that government contracts for the technology "include safety and security terms for data ownership, civil rights, civil liberties and privacy, adverse incident reporting and other key areas."
The bill would also mandate the appointment of chief AI officers responsible for procurement efforts – including the aforementioned audits, reviews and risk assessments. Establishing pilot programs to streamline AI adoption at the federal level is another requirement.
[7]America's War on Drugs and Crime will be AI powered, says Homeland Security boss
[8]Senate AI roadmap's piecemeal legislation is ideal, says former FTC tech chief
[9]Lawmakers advance bill to tighten White House grip on AI model exports
[10]US senators' AI roadmap aims for $32B in R&D spending
The bill has the support of groups including the Institute of Electrical and Electronics Engineers, the Center for Democracy and Technology, and the AI Procurement Lab.
"As agencies consider incorporating AI into government services and other processes, they must do so responsibly," Center for Democracy & Technology president and CEO Alexandra Reeve Givens observed. "The bipartisan PREPARED for AI Act lays a strong foundation by codifying transparency, risk evaluation, and other safeguards that will help agencies make smarter and more informed procurement decisions."
[11]
The US federal government has already enacted some AI laws, and president Biden issued an [12]executive order that requires safe use of the technology, plus rules that aim to ensure the Feds use a " [13]responsible " approach when putting it to work.
The PREPARED for AI Act appears to be the first attempt to tackle government procurement of such systems.
A Senate aide told The Register that chairman Peters hopes to have the legislation out of committee this summer. The aide didn't say whether a hearing date for the bill had been set – meaning it's unclear when or if it will be deemed ready for a vote. ®
Get our [14]Tech Resources
[1] https://www.congress.gov/bill/118th-congress/senate-bill/4495/text
[2] https://www.hsgac.senate.gov/media/dems/peters-and-tillis-introduce-bipartisan-bill-to-ensure-the-federal-government-safely-and-responsibly-purchases-and-uses-artificial-intelligence/
[3] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_onprem/publicsector&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=2&c=2ZmsXop2gE7zTFWOywMkkDAAAAFU&t=ct%3Dns%26unitnum%3D2%26raptor%3Dcondor%26pos%3Dtop%26test%3D0
[4] https://regmedia.co.uk/2024/06/12/prepared-for-ai-act-senate.pdf
[5] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_onprem/publicsector&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44ZmsXop2gE7zTFWOywMkkDAAAAFU&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[6] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_onprem/publicsector&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33ZmsXop2gE7zTFWOywMkkDAAAAFU&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[7] https://www.theregister.com/2024/05/07/dhs_ai_civil_liberties/
[8] https://www.theregister.com/2024/05/15/senate_ai_roadmaps_piecemeal_legislation/
[9] https://www.theregister.com/2024/05/23/us_lawmakers_advance_bill_to/
[10] https://www.theregister.com/2024/05/15/us_ai_roadmap_proposal/
[11] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_onprem/publicsector&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44ZmsXop2gE7zTFWOywMkkDAAAAFU&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[12] https://www.theregister.com/2024/05/27/government_agencies_slowly_getting_onboard/
[13] https://www.theregister.com/2023/05/25/white_house_ai/
[14] https://whitepapers.theregister.com/
Senators Gary Peters (D-MI) and Thom Tillis (R-NC) introduced the Promoting Responsible Evaluation and Procurement to Advance Readiness for Enterprise-wide Deployment (PREPARED) for AI [1]Act yesterday. The aim, [2]according to the pair, is to ensure the government can make use of the benefits of AI "while safeguarding against potential risks and harms."
"Artificial intelligence has the power to reshape how the federal government provides services to the American people for the better, but if left unchecked, it can pose serious risks," warned Peters. "These guardrails will help guide federal agencies' responsible adoption and use of AI tools, and ensure that systems paid for by taxpayers are being used safely and securely."
[3]
The [4]bill [PDF] proposes requiring government agencies to conduct a risk assessment exercise before acquiring and implementing AI. Risk audits would be required regularly after systems are implemented.
[5]
[6]
The bill also requires agencies to consider their AI implementation plans with regards to rights and safety of the public and requires that government contracts for the technology "include safety and security terms for data ownership, civil rights, civil liberties and privacy, adverse incident reporting and other key areas."
The bill would also mandate the appointment of chief AI officers responsible for procurement efforts – including the aforementioned audits, reviews and risk assessments. Establishing pilot programs to streamline AI adoption at the federal level is another requirement.
[7]America's War on Drugs and Crime will be AI powered, says Homeland Security boss
[8]Senate AI roadmap's piecemeal legislation is ideal, says former FTC tech chief
[9]Lawmakers advance bill to tighten White House grip on AI model exports
[10]US senators' AI roadmap aims for $32B in R&D spending
The bill has the support of groups including the Institute of Electrical and Electronics Engineers, the Center for Democracy and Technology, and the AI Procurement Lab.
"As agencies consider incorporating AI into government services and other processes, they must do so responsibly," Center for Democracy & Technology president and CEO Alexandra Reeve Givens observed. "The bipartisan PREPARED for AI Act lays a strong foundation by codifying transparency, risk evaluation, and other safeguards that will help agencies make smarter and more informed procurement decisions."
[11]
The US federal government has already enacted some AI laws, and president Biden issued an [12]executive order that requires safe use of the technology, plus rules that aim to ensure the Feds use a " [13]responsible " approach when putting it to work.
The PREPARED for AI Act appears to be the first attempt to tackle government procurement of such systems.
A Senate aide told The Register that chairman Peters hopes to have the legislation out of committee this summer. The aide didn't say whether a hearing date for the bill had been set – meaning it's unclear when or if it will be deemed ready for a vote. ®
Get our [14]Tech Resources
[1] https://www.congress.gov/bill/118th-congress/senate-bill/4495/text
[2] https://www.hsgac.senate.gov/media/dems/peters-and-tillis-introduce-bipartisan-bill-to-ensure-the-federal-government-safely-and-responsibly-purchases-and-uses-artificial-intelligence/
[3] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_onprem/publicsector&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=2&c=2ZmsXop2gE7zTFWOywMkkDAAAAFU&t=ct%3Dns%26unitnum%3D2%26raptor%3Dcondor%26pos%3Dtop%26test%3D0
[4] https://regmedia.co.uk/2024/06/12/prepared-for-ai-act-senate.pdf
[5] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_onprem/publicsector&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44ZmsXop2gE7zTFWOywMkkDAAAAFU&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[6] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_onprem/publicsector&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33ZmsXop2gE7zTFWOywMkkDAAAAFU&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[7] https://www.theregister.com/2024/05/07/dhs_ai_civil_liberties/
[8] https://www.theregister.com/2024/05/15/senate_ai_roadmaps_piecemeal_legislation/
[9] https://www.theregister.com/2024/05/23/us_lawmakers_advance_bill_to/
[10] https://www.theregister.com/2024/05/15/us_ai_roadmap_proposal/
[11] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_onprem/publicsector&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44ZmsXop2gE7zTFWOywMkkDAAAAFU&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[12] https://www.theregister.com/2024/05/27/government_agencies_slowly_getting_onboard/
[13] https://www.theregister.com/2023/05/25/white_house_ai/
[14] https://whitepapers.theregister.com/
The actual referenced definition of AI in the bill is 15 USC 9401:
The term "artificial intelligence" means a machine-based system that can, for a given set of human-defined objectives, make predictions, recommendations or decisions influencing real or virtual environments. Artificial intelligence systems use machine and human-based inputs to-
(A) perceive real and virtual environments;
(B) abstract such perceptions into models through analysis in an automated manner; and
(C) use model inference to formulate options for information or action.
From my perspective, almost every bit of code I work with meets that definition. From your spam filter to your image editor (e.g. identify which photos have faces in them, "smart" editing brushes, automatic exposure adjustment) to even your automatic lights that turn off if you're too still for too long. I see it as too broad a definition, by far.