CenDemTech

ICYMI: Earlier this week, CDT’s Aliya Bhatia wrote in The Guardian about the growing wave of age verification laws that could force people to upload government IDs just to access everyday online platforms. From privacy to free expression, the piece explores what’s at stake as govts push new restrictions online. Read more: https://www.theguardian.com/commentisfree/2026/may/12/website-age-verification-government-identification

In WaPo’s AI & Tech Brief newsletter: “According to a new report from the AI Governance Lab at CDT and MIT, fine-tuning models can have unpredictable consequences on AI models’ safety guardrails — in ways that even the most sophisticated AI researchers don’t completely understand… seemingly minor changes to model ‘weights’ can have major downstream effects.” https://s2.washingtonpost.com/camp-rw/?trackId=602d3a45ae7e8a31ba8ea23c&s=69f8ecbbe9f14b167bc6c4ce&utm_medium=email&utm_source=newsletter&utm_campaign=wpi_ai_tech_brief&linknum=5&linktot=39

“Another study, called ‘Hand in Hand,’ done by the Center for Democracy and Technology, has shown that 72% of parents, 71% of teachers, and 64% of students have all reported that they felt AI was diminishing their critical skills.” https://www.theteenmagazine.com/ai-is-the-1-killer-of-common-sense

CDT's Hannah Quay-de la Vallee and Quinn Anex-Ries break down how the Pentagon's dispute with Anthropic isn't just a defense issue. It's disrupting AI adoption across every level of government, from federal agencies to state, local, and tribal governments. https://cdt.org/insights/chain-reaction-what-the-pentagon-anthropic-dispute-means-for-civilian-agencies-across-all-levels-of-government/

Happening today at 4 PM ET: Join the Educating All Learners Alliance and New America for a cross-sector briefing on “Prioritizing Students with Disabilities in AI Policy.” Hear from researchers, advocates, and industry leaders, including CDT’s Ariana Aboulafia, on building AI tools that work for every student from day one. https://cdt.org/event/a-cross-sector-briefing-prioritizing-students-with-disabilities-in-ai-policy/

How should the government protect your personal data, and who's holding agencies accountable when they don't? Join CDT June 1 (12–1pm ET) for a virtual panel on government data privacy with experts from civil society and government agencies. Register now: https://cdt-org.zoom.us/webinar/register/WN_XdkXdPS8SUGyvO4VWKIPxQ

CDT's Ariana Aboulafia is speaking this Friday (5/8) at a cross-sector briefing hosted by the Educating All Learners Alliance + New America on the launch of "Prioritizing Students with Disabilities in AI Policy." Join 4 PM ET to hear how we move from reactive compliance to proactive design: https://cdt.org/event/a-cross-sector-briefing-prioritizing-students-with-disabilities-in-ai-policy/

“...Elizabeth Laird, director of equity in civic technology at CDT, described the education sector as a ‘target-rich environment,’ as schools and vendors collect highly sensitive information about tens of millions of Americans, including minors.” https://www.govtech.com/education/instructure-makes-deal-with-hackers-as-experts-talk-data-security

From the 1960s’ “Art by Telephone” to generative AI, expression has always moved through intermediaries. The First Amendment question isn’t whether machines “speak,” but whether people can speak, & inquire, through them. Read the blog by CDT’s Becca Branum: https://cdt.org/insights/art-by-telephone-art-by-algorithm-expression-ai-and-the-first-amendment/

“A 2025 report by CDT also shows that an average of 71% of K-12 teachers reported that when students use AI to complete their schoolwork, it is hard for the teachers to understand whether student work is their own.” https://www.mississippifreepress.org/opinion-using-ai-in-schools-may-do-more-harm-than-good/

CDT joined a coalition of organizations in a letter to Rep. Ted Lieu (D-CA) urging him to ensure his bipartisan American Leadership in AI Act is not packaged with or used as cover for broader efforts to preempt state artificial intelligence laws. https://cdt.org/insights/cdt-joined-a-letter-to-representative-ted-lieu-urging-him-to-ensure-his-recent-bipartisan-american-leadership-in-ai-act-is-not-tied-to-broader-attempts-to-preempt-state-ai-laws/

NEW: Voter registration databases are secure, and the claims saying otherwise are misleading. CDT's Geoff Hale, plus Brennan Center's Katie Lurie & Veronica Degraffenreid break down what the security record actually shows. https://cdt.org/insights/op-ed-voter-registration-database-protections-are-resilient/

CDT’s Travis Hall in Governing: “[Colorado’s SB 189 bill is] no longer really a discrimination bill. It is now almost entirely a disclosure-focused bill.” https://www.governing.com/policy/colorado-revamps-landmark-ai-law

CDT’s AI Governance Lab is driving robust, technically-informed solutions for effective AI regulation. Our team centers the rights and interests of historically marginalized communities to ensure AI policy advances equity and democratic values. Find out more: https://cdt.org/cdt-ai-governance-lab/

As the midterms approach, political influencers are playing a bigger role than ever in how voters get information, engage with campaigns, and shape public opinion online. Analysis from CDT explores what this means for transparency, accountability, and democratic norms. https://cdt.org/insights/political-influencers-and-democracy-in-the-digital-age/