A Focus on AI Innovation and Safety

AI is being widely used—by both responsible parties and bad actors—even as public policies and best practices for the technology are being developed. This week we want to call attention to three blogs on key AI issues:

Combating abusive AI-generated content—Deepfakes generated by AI can be used to undermine elections, commit financial fraud, and cyberbully. This Microsoft on the Issues blog identifies six ways to combat this threat, including by developing new laws to protect personal privacy.

Navigating cyberthreats and strengthening defenses in the era of AI—This Microsoft Security blog discusses research on the use of LLMs in cyberattacks—and the development of countermeasures.

The need for the National Artificial Intelligence Research Resource (NAIRR)—This VFI blog highlights a pilot program to expand AI access for critical academic and scientific research. But further federal support is needed to expand the program.

Thank you for reading!

This Week in Washington 

  • The Daily Yonder: Despite being eligible to apply for the Affordable Broadband Program (ACP), an estimated 5.4 million rural households did not apply to the program due to a lack of awareness and proximity to resources like local libraries and other anchor institutions that could have helped facilitate enrollment.
  • POLITICO and The Hill: President Joe Biden became the first Democratic presidential candidate to embrace TikTok as part of his campaign strategy to reach young voters, showcasing the platform’s steady persistence in Washington. Although this is a win for TikTok, President Biden hasn’t changed his official view on the app, and lawmakers are continuing to urge the Department of Commerce to add its parent company, ByteDance, to its export control list due to concerns over U.S. user data.
  • The Hill: Senator Chuck Schumer (D-NY) announced support for a newly updated version of the Kids Online Safety Act (KOSA), while several LGTBQ advocacy organizations including GLAAD and the Human Rights Campaign dropped their opposition to the bill after its sponsors changed the text to address concerns they had raised last year. The bill now has over 60 Senate co-sponsors, and could pass there, although its counterpart in the House has a less-certain future. 
  • The Washington Post: One Democratic member of the Federal Trade Commission spoke out against proposed restrictions to limit kids’ access to social media sites, believing that they won’t work. Commissioner Alvaro Bedoya suggested that young people would be able to work around restrictions, and many restrictions would violate Constitutional rights, aligning with concerns from tech industry groups and digital rights advocates.
  • Nextgov/FCW: Even as lawmakers look toward regulating AI, some are also pushing for the internal use of the technology, believing that it could help improve Capitol Hill operations and inform policy over time.
  • FedScoop: A new, bipartisan proposal in the U.S. Senate, the AI and Critical Technology Workforce Framework Act, would direct the National Institute of Standards and Technology (NIST) to create a workforce framework for AI and assess whether other critical or emerging technology areas might also benefit from frameworks. The bill’s authors want to build upon NIST’s existing National Initiative for Cybersecurity Education (NICE) framework that is intended to help employers build their cyber workforces.
  • Nextgov/FCW: A newly published Federal Communications Commission (FCC) rule, set to take effect in March after being adopted by the FCC late last year, will require telecommunications and voice over IP providers to notify the FCC, Secret Service, and FBI of data breaches within seven business days of discovery. Companies must also notify customers about data breaches “without unreasonable delay” within 30 days of determining a breach has occurred, unless law enforcement requests a delay.
  • CyberScoop: The Cybersecurity and Infrastructure Security Agency (CISA) released the 2024 priorities for the Joint Cyber Defense Collaborative, including defending against advanced persistent threat (APT) operations, raising baseline protections for critical infrastructure owners and operators, and anticipating emerging technology and risks.

Article Summary

  • Texas Tribune: Texas, where 7 million residents lack access to high-speed internet connections, will receive more than $3.3 billion in federal money to fund a dramatic expansion of broadband. ISPs and residents are hopeful that this effort will help many get connected, though the state office responsible for disbursing the funds is still hiring staff to oversee the work.
  • Cardinal News: Lawmakers in both chambers of Virginia’s state legislature passed bills this week to help address conflicts between ISPs and utilities over federal broadband infrastructure funding. The new legislation’s sponsors hope the bill will speed the process up and assist the state in meeting a federal deadline for spending about $750 million in pandemic relief funds before the end of 2026.
  • The Hill: Meta announced their Instagram and Threads products will no longer be proactively promoting political content to users in their feeds unless users change their settings to allow it.
  • Axios and Business Insider: While older generations have expressed concerns over how AI will affect their jobs, Gen-Z, also known as “AI-natives,” are embracing the technology, believing that digging into it will give them a career edge. Meanwhile, the University of Pennsylvania will be the first Ivy League school to offer an AI-focused undergraduate degree: a Bachelor of Science in Engineering in Artificial Intelligence (AI).
  • Hawaii News Now: A state lawmaker in Hawaii has authored legislation, intended to combat the proliferation of “deepfakes,” to require that state campaigns label any use of artificial intelligence in campaign materials starting 90 days before any primary or general election. Hawaii would join a number of other states working to restrict or label the use of AI, particularly video and image generation, in political campaigns.

Featured Podcast


  • Tools and Weapons with Brad Smith
    Yves Ubelmann is a technologist, artist, and architect who is on a mission to digitally preserve the world’s cultural and natural heritage. He is the founder and CEO of Iconem, a company that creates stunning 3D models of endangered sites and environments. In this episode, Brad and Yves take a boat trip through Venice, touring the subject of his most ambitious project: an AI-generated digital twin of one of the world’s most magnificent cities. Inspired by his grandfather’s work to restore war-damaged heritage sites in France, Yves shares how he is using AI to memorialize the world’s heritage in 3D and raising awareness of the effects of climate change on our planet. (Yves Ubelmann: Saving our heritage, one digital twin at a time – February 8, 2024)