To view this email online, paste this link into your browser:
[link removed]





AUGUST NEWSLETTER  

([link removed])

Real-World Usage Data is Crucial to Better AI Policy

As companies and governments rapidly integrate AI into new systems and domains, policymakers are scrambling to regulate AI in order to mitigate risks and maximize potential benefits. To make good policy, policymakers need to better understand how AI technologies are actually used, but as we make clear in a new CDT Research paper ([link removed]), “Grounding AI Policy: Towards Researcher Access to AI Usage Data,” this evidence is hard to come by.

([link removed])

 Illustration for a CDT Research report, entitled “Grounding AI Policy.” Two silhouettes are observing an enclosed digital “plant.” Their ability to observe data on the plant informs their actions on it.

Interactions with AI systems are usually not publicly observable, and AI companies are reluctant to share data about those interactions even with vetted researchers. While many companies do allow researchers and other external parties to probe their systems for vulnerabilities and harmful errors, those methods only reveal the worst use cases, and don’t offer empirical insights about the real-world harms users experience.

In the paper, we describe this information gap, why it should be closed, and what challenges there are to doing so. We identify three potential methods that can help address this information gap — data donations, transparency reports, and direct access to information about users’ interactions with products — and the benefits and challenges presented by each one. Finally, we offer recommendations for how AI companies and lawmakers can implement these methods in ways that benefit researchers and ultimately the public, while safeguarding users’ privacy.

In Case You Missed It

— In a CDT report, “ To Reduce Disability Bias in Technology, Start With Disability Data ([link removed]),” we explain that understanding and mitigating the problems with disability-related data is core to tackling technology-facilitated disability discrimination — particularly in several high-stakes contexts like employment, benefits, and healthcare — and to improving the lives of people with disabilities overall.

([link removed])

Cover graphic for the 2023 CDT annual report, entitled "Towards a Responsible Digital Future." Image of an open atrium, looking up.

— CDT released recommendations for how developers of generative AI technologies can help protect election integrity ([link removed]) by preventing the creation of harmful content at the source, including through service usage policies, product interventions, policy enforcement, and transparency.

— In a blog post, we discuss what the Supreme Court’s decision in Moody v. NetChoice says ([link removed]) about how governments may appropriately impose transparency requirements.

— The Biden-Harris Administration published a report with recommendations on how to keep youth safe online ([link removed]), which aligned with some of the principles and strategies we’ve outlined to enhance and strengthen online safety for young users, including the necessity of engaging youth in policymaking and development of intervention strategies.

— The 2023 CDT Annual Report, “Towards a Responsible Digital Future ([link removed]),” is available on CDT’s website.

CDT in the Press 

— CDT’s Dhanaraj Thakur spoke with Wired ([link removed]) about patterns of disinformation against women of color political candidates: “As a society we do not treat all politicians the same because of race and gender and other factors,” he said. “But platforms want to treat everyone the same.”

Screenshot of CDT's Elizabeth Laird, a woman with brown hair wearing a black shirt with a kitchen in the background, appearing on a Cox Media TV segment.

— CDT’s Matt Scherer discussed trends in state AI policymaking with the Washington Post ([link removed]): “California has outsize influence on tech policy. Whatever bills pass this session in California are going to be very impactful nationwide,” he said. “One state can be an aberration. Two states is a trend.”

— CDT’s Elizabeth Laird joined Cox Media Group ([link removed]) for a segment on AI use in the classroom: “Schools and teachers are grappling with this as we speak. We did find that schools have more policies in place than they had last year and they are starting to provide training, but the actual substance of it has a long way to go,” she said.

Graphic with thought bubble with patriotic decorations and text: "Future of Speech Online (FOSO): AI, Elections, & Speech. September 16-17, 2024."

CDT "in Person"

— Join CDT and Stand Together Trust on September 16-17 for this year’s Future of Speech Online ([link removed]) conference! Hear from legal experts, technologists, and election experts as we discuss the legal frameworks and technical infrastructure needed to bolster free expression and protect free and fair elections globally.

— We are thrilled to announce CDT’s 2024 Tech Prom, which will take place on November 14 at the Anthem in Washington, D.C. Learn more about the event, including how to sponsor or purchase tickets, on our Tech Prom page ([link removed]). 

Laura Lazaro Cabrera, smiling in a light colored top in front of a CDT Europe backdrop.

Staff Spotlight
Laura Lazaro Cabrera ([link removed]), Counsel and Director of the Equity and Data Programme, CDT Europe

How long have you been working in digital rights? I have been working in digital rights since 2020, when I immersed myself in the universe of data protection and new technologies.

What is your proudest moment while here at CDT? Convincing the European Commission that they should invite civil society to participate in the development of rules that will determine how general purpose AI model providers are regulated in the EU.

What is the best book you've read recently?? It's a tie between Cien cuyes by Gustavo Rodríguez, and Los Genios by Jaime Bayly.

Cats or dogs? Dogs!

#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]