The House Task Force on AI

The House created a task force on Artificial Intelligence last year which was tasked with developing a US vision for AI adoption, innovation and governance. They provided 66 key findings and 85 recommendations over 15 different issue areas.

The Steptoe law firm summarized the key aspects of this report (link below). Congress has some interesting challenges in implementing this.

The high level principles are motherhood and apple pie, but it doesn’t hurt to include them. They include identifying AI issue novelty, promoting AI innovation, protecting against AI risks and harms, taking an incremental approach and other principles.

I am not going to repeat their recommendations – that would make for a really long blog post, but I recommend checking out the entire Steptoe blog. I think you will find it interesting. Here are some key points:

A. Government use

The government is already using AI. The government should support flexible governance and use AI to improve efficiency, adopt standards, improve security and encourage data governance strategies.

B. Federal Preemption of State Law

Wisely, they kicked this can down the road. Likely they had an opinion – my guess is in favor of more federal power – but that would be politically unpopular, especially for the party in charge. As a result, they said they should study it.

C. Data Privacy

They say that Congress should provide access to your data in privacy-enhanced ways. My opinion is that this is a really hard problem.

D. National security

Congress should focus on overseeing AI activities for national security, expand AI training at the DoD, oversee autonomous weapons policies and support international cooperation.

E. Research, Development and Standards

Congress should continue to fund federal R&D efforts and bolster US standards efforts.

F. Civil Rights and Liberties

AI outputs should have human intervention when AI is used in highly consequential decisions. Congress should support transparency and training in AI decision making to mitigate discrimination.

G: Education and workforce

Congress should fund AI education in K-12 schools, colleges and workforce development.

H. Intellectual property

Congress should clarify intellectual property laws and regulations and figure out how to counter harm from deepfakes.

I. Content authenticity

Congress should support technical methods to identify fake content and ensure victims have ways to counter harm from the fake content.

J. Open and closed systems

Congress should monitor risks from open-source models and encourage innovation.

K. Energy use and data centers

Support and increase federal investments in research that enables innovations in AI; strengthen efforts to track AI power usage and creates new energy use metrics.

L. Small businesses

Congress should support AI literacy and adoption for small businesses and reduce their compliance burdens.

M. Agriculture

Congress should identify opportunities for advancing AI in precision agriculture, specialty crops and land management and review risks posed by AI in financial markets.

N. Healthcare

Congress should ensure that healthcare AI is safe, transparent and effective and support the development of standards for liability related to AI issues.

O. Financial services

Congress should ensure protections are in place in the financial services and housing sectors by encouraging regulators to increase their expertise in AI.

This is a very brief synopsis of what the task force report recommended. Even at this high level, Congress has a lot of work to do in order to implement these recommendations. We will see if they accomplish them.

Credit: Steptoe

The full House report can be downloaded here.

by