Blog

Arizona, Florida, Nebraska, Ohio and Wisconsin joined the FTC’s complaint against telehealth provider Hims & Hers Health, Inc.  The complaint alleges that company shared consumers’ health information with Meta, Snap, and other third parties, in violation of state consumer protection laws, FTC Act Section 5, and ROSCA.  (Case timeline here)

Sen. Markey (D- MA) sent a letter to Google asking it to allow schools to disable AI features in search. Google has said that it cannot disable tools like A.I. Overview and A.I. Mode because they are too “deeply integrated” into search.

Markey’s letter follows from Times reporting on the severe impact of A.I. chatbots on students’ mental health.

Relatedly, New York City school districts are in a “standoff” with Google, after Google announced that schools that banned AI would also lose access to Google search. Coverage here () and here

Also related to the linkage of Google search and AI overviews, a German judge has reasoned that AI is not necessary to search the internet, as part of a decision ruling that Google is liable for false statements in AI Overviews. 

AI law firm Crosby purchased Malpractice Insurance for its A.I. Agents. Crosby secured $5 million of A.I. liability insurance from Testudo, a two-year-old insurance company with underwriting capacity backed by Lloyd’s of London. Testudo is part of a group of new insurers like Harbor and the Artificial Intelligence Underwriting Company that are positioning themselves to fill coverage gaps around AI implementation.

(compiled by Elyse Cox)

The Citizen Lab reported on how targeted ads could be used for zero-click spyware.

The EFF has publicized a report noting various issues with Medicare’s AI automated decision making.

Wired has reported the ways DHS and ICE have used their expanded budget to go after left-wing groups.

A New Mexico judge has ordered $25 billion in civil penalties following the recent jury verdict in the state privacy suit.

A Federal Judge in the Northern District of Oklahoma has ruled that “a large-scale, dragnet-type surveillance system like the [Automated License Plate Reader] technology” such as Flock can be a search. (news article discussing it; opinion at 28-38)

A draft bill from Anthropic related to AI has been leaked; it has some discussion of privacy.

The Supreme Court heard argument in Salazar v. Paramount Global, a case about the scope of the Video Privacy Protection Act (transcript here)

Eight states have sued openAI in coordination independently in state courts, claiming danger to youth.

It appears that Crosby (an alleged AI law firm that has taken out many ads on the MTA) has over the summer taken out liability insurance for its agents.

California has restricted CIPA’s private right of action under Section 638.51 to solely AG enforcement.

A Representative, Darrell Issa (R-CA), has introduced a bill targeting VPNs and foreign websites accused of copyright infringement.

OpenAI has responded to an EU directive requiring text provenance by allowing users to opt in to text watermarking (full paper here).

A New Mexico jury found Facebook liable for violations of the New Mexico unfair competition law for misstatements regarding its privacy policies.

Meta’s AI reportedly gave a user’s address to a buyer on Facebook Marketplace without authorization (or, indeed, even notifying the user that it would do so).

Trump has talked up AI firm’s self-policing.

Stanford used AI to alter the appearance of certain students, and even swapped out a student for a different student of a different race.

The Supreme Court issued a stay in DHS v. League of Women Voters regarding the scope of the Systematic Alien Verification for Entitlements (SAVE) program.

Welcome back to a new year of PRG!

Many recent articles have shows issues with flock surveillance cameras and the way they are being used.

The European Commission has stated it is moving forward with a plan to draft legislation on limiting social media to minors under 13.

As AI models continue to develop—and get into high-profile cybersecurity issues—some AI insiders have begun to publicly focus more on safety. Others insiders have split on what they believe to be the best approach between regulation, industry-lead safeguards, and reliance on individual companies.

Surveillance firm Palantir has faced heavy scrutiny for its ties to the Trump Administration, but now even Democrats are getting flak for connections to the firm. Democratic challenger Colin Allred has accused Representative Julie Johnson of profiting from Palantir’s activities, having held stock in the company while serving in Congress. The company’s collaborations with ICE have proven particularly caustic, and other lawmakers have quietly cut ties with the firm, including both Democrats and Republicans. While Palantir affiliates have tried to push back on its negative reputation, Palantir is quickly becoming a liability for political candidates. 

Judge F. Dennis Saylor IV has extended the deadline for public colleges and universities to submit data on the race & gender of admitted students, which the Trump Administration demanded from the schools by March 18th. Judge Saylor noted ongoing personnel and infrastructure issues in the Department of Education, which would likely put student data at risk, and raised the possibility of a preliminary injunction against the order. Nevertheless, the states’ contention that the data collection was politically motivated did not persuade the judge, who noted the long history of similar data collections in association with federal anti-discrimination laws. 

Social media platform Reddit is considering implementing a verification program, in which users would be asked to submit face photographs or biometric data to verify their humanity. Reddit is one of the largest social media sites on the internet and has long allowed pseudonymous accounts. Still, the flood of AI-generated content is presenting major problems for users and administrators alike. The stakes are high as the company’s leadership seeks to maintain the human composition of Reddit’s userbase while avoiding scaring off privacy-conscious users. 

The Senate recently passed the Children and Teens’ Online Privacy Protection Act, or COPPA 2.0, with unanimous support, pushing forward “the most significant update to our nation’s online privacy protections for young people” in over 25 years. While the bill still needs approval from the House, COPPA 2.0 explicitly “cannot be read to require” age verification. Age verification poses major constitutional and political questions, to say nothing of the difficulties in implementing a solution. The strategy may be off the table for now, but various states and developed countries have implemented age verification with mixed results, even where the public has strongly supported the measures.

(compiled by David Gonzalez)

The Electronic Frontier Foundation analyzed the SAFE Act, a new legislative proposal to reform and reauthorize federal surveillance powers under national security law. 

Meta is facing a class-action lawsuit over privacy concerns with its AI smart glasses. 

A former DOGE employee is accused of retaining sensitive Social Security information about more than 500 million living and deceased Americans for commercial use.

AI-assisted coding tools were linked to production outages at Meta this week. 

From the PRG community: 

Nina Loshkajian, Technology & Racial Justice Collaborative Fellow at the NYU Law Center on Race, Inequality and the Law, argues for renewed attention to disparate impact protections in the age of AI. 

Upcoming events: 

Tomorrow, March 12, the Innovation Policy Colloquium will host Feng Fu, Associate Professor of Mathematics at Dartmouth, for a presentation on steering collective behavior in hybrid human-AI systems.

(compiled by Nala Sharadjaya)