From Center for Democracy & Technology <[email protected]>
Subject CDT Weighs in on U.S. Government Progress on AI EO
Date February 23, 2024 4:20 PM
  Links have been removed from this email. Learn more in the FAQ.
  Links have been removed from this email. Learn more in the FAQ.
To view this email online, paste this link into your browser:
[link removed]





FEBRUARY NEWSLETTER  

([link removed])

CDT Weighs in on U.S. Government Progress on AI EO

In October of 2023, the Biden Administration issued its Executive Order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence. CDT welcomed the Executive Order ([link removed]) (EO), which lays out a roadmap for action by various parts of the federal government to ensure responsible AI practices by both the private and public sectors. We also urged the Biden Administration to move quickly to implement the EO and ensure its effective role in driving meaningful change. 

Four months on from the EO’s release, agencies are stepping up to the plate. In November 2023, the U.S. Office of Management and Budget (OMB) released draft guidance for federal agencies’ use of AI, which will ultimately help ensure that government AI practices respect rights and don’t interfere with access to equal opportunities or critical services. In December, CDT submitted comments noting the ways in which OMB should strengthen that guidance ([link removed]), with particular emphasis on strengthening transparency requirements and narrowing agencies’ ability to waive the rules. In January, CDT joined with several other civil society organizations to highlight key changes the group agreed OMB should make to the guidance ([link removed]).

The National Institute for Standards and Technology (NIST) also requested public feedback on how it should carry out several of its responsibilities under the EO ([link removed]), which include developing frameworks for evaluating and auditing the capabilities of various AI technologies. In response, CDT urged NIST ([link removed]) to govern generative AI no differently than it would AI in general, and to focus on nailing down a common set of rigorous and scientifically grounded concepts, approaches, and infrastructure for AI risk management. 

This month, CDT joined as a member of NIST’s AI Safety Institute Consortium ([link removed]), with CDT’s AI Governance Lab Director Miranda Bogen speaking at the Institute’s launch.

In another area, the EO directed the Department of Justice (DOJ) Civil Rights Division to coordinate with federal government agencies on enforcing civil rights and anti-discrimination laws in cases related to AI. CDT joined several civil society partners in providing recommendations to DOJ ([link removed]) on how the agency might accomplish that goal, specifically by developing guidelines on testing for algorithmic discrimination and strengthening the federal government’s enforcement of civil rights and consumer protection laws to address private sector entities’ use of AI. 

CDT’s work on AI Executive Order implementation will continue this spring, as we prepare to respond to NTIA’s new Request for Comments on Dual Use Foundation Models with Widely Available Model Weights ([link removed]) (the “open source” proceeding), as well as engaging with the Department of Labor, Department of Education, Department of Health & Human Services, and national security authorities on their various actions in the EO. With leadership from across our policy teams, our focus is on supporting the development and deployment of safe, trustworthy, and responsible AI.

In Case You Missed It

— CDT is launching a new project focused on content moderation policies and measures in the Global South ([link removed]), rooted in prior CDT work examining the ways tech companies design models to analyze text across languages. Given that the vast majority of tech company efforts to expand their content moderation systems have been geared toward English-language content, CDT’s research will critically examine how content moderation systems operate in non-English contexts. 

— CDT joined the Electronic Frontier Foundation (EFF) in filing an amicus brief in NetChoice v. Bonta ([link removed]), a case in which industry association NetChoice is challenging the California Age-Appropriate Design Code (AADC) for violating its members’ constitutional rights by forcing them to determine the age of their users and impede access to constitutionally protected speech. CDT and EFF argue that the AADC’s age-verification and censorship requirements are overbroad and unconstitutional, but that the law’s privacy protections are not.

— CDT joined a group of security experts in highlighting the detrimental impacts that several amendments to the UK Investigatory Powers Act would have ([link removed]) on digital security and privacy, namely that they would establish centralized control over crucial security updates and technological development and maintenance, which would significantly erode trust in digital services and compromise user safety.

— CDT submitted comments in support of an FTC rulemaking ([link removed]) to promote the right to repair — the right for consumers to have their electronic equipment serviced by the repair service provider of their choosing, and the right for independent repair service providers to compete to offer that choice.

CDT in the Press

— Regarding delays to a vote by Congress on reauthorizing controversial surveillance authority Section 702, CDT’s Jake Laperruque told Reuters ([link removed]), "It's time to stop punting and bring the debate over warrantless FISA surveillance to the House floor."

([link removed])

Screenshot of CDT President and CEO Alexandra Givens speaking with Cox Media's Kirstin Garriss.

— CDT CEO Alexandra Givens spoke with COX Media ([link removed]) about the New Hampshire robocalls that faked President Biden’s voice to convince voters that they did not need to vote in the state primary: “So, the second a voter gets a call saying that they don’t have to turn up on Election Day — where do they go to actually get the gut check on what to believe? [O]ur local election officials really have to be doing a good job of getting out authoritative information to reassure people.”

— CDT’s Aliya Bhatia talked with Contexte ([link removed]), discussing the problems with large language models backed by governments: "[These models may be] likely to adopt a worldview where the government is wholly positive and leave behind dissenting viewpoints… There is a real risk of government-backed models instilling a revisionist view of history and undermining democratic values." 

([link removed])

Illustration for CDT’s 2024 Spring Fling event. Blue and rose-colored text surrounded by light pink blossoms and orange bulbs.

CDT "in Person"

— CDT is pleased to announce our 2024 Spring Fling, a celebration that will take place during IAPP’s Global Privacy Summit. Join us on Tuesday, April 2 ([link removed]), from 8–10:30 PM at Dirty Habit in Washington, D.C. — you won’t want to miss it!

— On February 29, join CDT’s Ariana Aboulafia in a virtual panel at Mobility Matters 2024 ([link removed]), where she will discuss navigating AI as a tool for exploration and inclusion. 

Partner Spotlight

CDT is proud to have worked with the American Civil Liberties Union (ACLU) to advocate to the federal government to update guidance and increase enforcement against AI harms extending across employment, housing, public benefits, and other critical areas. Most recently, we worked with the ACLU and several partners to identify some shared priorities for the Office of Management and Budget’s guidance on agency use of AI ([link removed]) required under the AI EO.

The ACLU dares to create a more perfect union — beyond one person, party, or side. Their mission is to realize this promise of the United States Constitution for all and expand the reach of its guarantees. You can learn more about the ACLU and their work on their website ([link removed]).

Sarah Zolad, wearing glasses, a white collared shirt and a black sweater, smiling in front of a brick outdoor setting.

Staff Spotlight
Sarah Zolad ([link removed]), Director of Finance and Staff Operations

How long have you been working in digital rights? I have spent my career working in operations in the nonprofit space, particularly in choosing, building, and instituting many different technology solutions. There was always a part of any project where I needed to safeguard information - such as personal data or health information. It has been a throughline connecting my career. I am excited to have this opportunity to work more broadly on a systemic level.

What is your proudest moment while here at CDT? I have only been here for a few months, but I'd have to say I am proudest of how I have been able to merge all of my past experience to ensure that business operations run smoothly. It has been delightful to get to know the smart, supportive, and dedicated staff and collaborate with them on diverse projects.

What is your fandom? I love a good Cozy Mystery, either in book or TV form.  I find it so calming to know that at the end of the story the hero will save the day, the villain will be chastened and apologetic, and usually there will be a collection of interesting and far-fetched characters. It is truly amazing that anyone is still alive in quaint villages anywhere in the world.

Dogs or cats? Unfortunately, due to allergies, I can no longer have either. But I am very happy with many plants of all sorts of shapes and sizes. I love variegated leaves and plants that do the unexpected. Currently I have a black hellebore blooming in my yard. They come in purple, green, and black, and bloom in January. Every morning, I can see it as I leave the house and it reminds me that spring is just around the corner.

#CONNECT WITH CDT

SUPPORT OUR WORK ([link removed])

([link removed])

([link removed])

([link removed])

([link removed])





1401 K St NW Suite 200 | Washington, DC xxxxxx United States

This email was sent to [email protected].
To ensure that you continue receiving our emails,
please add us to your address book or safe list.

manage your preferences ([link removed])
opt out ([link removed]) using TrueRemove(r).

Got this as a forward? Sign up ([link removed]) to receive our future emails.
email powered by Emma(R)
[link removed]
Screenshot of the email generated on import

Message Analysis