The letters

The open letters sent to the 150 largest American school districts. Each letter is followed by its own footnotes.

  1. November 20, 2026

    Follow-up letter

    Reports new developments on YouTube and Gemini AI, responds to a critique from Lightspeed Systems, and urges the districts to stand up to Google.

  2. October 5, 2026

    First letter

    Asks the districts to disable YouTube and Gemini AI on school-issued devices and to investigate student internet use.

Signatories

Who signed the letters

  • The AI Psychological Research Coalition
  • Becca Schmill Foundation
  • Buckets Over Bullying
  • Center for Countering Digital Harm
  • The Center for Responsible Technology
  • The Child First Policy Center
  • Digital Childhood Institute
  • Distraction-Free Schools Policy Project
  • DMV Unplugged
  • Fairplay
  • Four Norms
  • Institute for Families & Technology
  • Institute for Family Studies
  • Protect Young Eyes
  • Schools Beyond Screens
  • Scrolling 2 Death
  • Smartphone Free Childhood U.S.
  • The Screentime Consultant Emily Cherkin
  • LME Global Dr. Jared C. Horvath

Letter 2 · Follow-up

Follow-up letter, November 20, 2026

November 20, 2026

Re: Follow-Up to Open Letter to School Districts, District Leadership, and School Boards

Dear School Districts, School District Leaders, and School Boards (the “Districts”):

On October 5, 2026, thirteen leading organizations and experts on technology in schools wrote an Open Letter to the Districts, (i) informing the Districts of recent litigation developments and reporting concerning student use of YouTube, Gemini AI features, and the internet on school-issued devices; (ii) requesting that the Districts disable YouTube and Gemini AI features on school-issued devices, and (iii) requesting that the Districts investigate student internet use on school-issued devices, disclose the results to the public, and reform their policies to ensure student safety.1

Now, nineteen leading organizations and experts – the thirteen who joined the Open Letter, plus six more2 – write the Districts to:

  1. Inform the Districts of additional reporting and litigation that reinforce the points made in the Open Letter and demonstrate the safety and legal risks of YouTube, Gemini AI, and student internet access on school-issued devices;
  2. Respond to a critique of the Open Letter from Lightspeed Systems, which apparently has done business with Google; and
  3. Urge the Districts to stand up to Google, which has recently sought to bully the Districts into forcing Gemini AI features onto students by threatening to take away Google Search functionality if the Districts refuse to do so.

I. Additional Reporting and Litigation Developments Related to YouTube, Gemini AI, and Student Internet Use on School-Issued Devices.

a. An NBC Report Reinforces the Dangers of YouTube and Supports the Open Letter’s request that the Districts Turn Off Student Access to YouTube.

On October 20, 2026, NBC News reporter Tyler Kingkade issued a report, Hooked on YouTube, further demonstrating the shocking extent to which students in the Districts are drawn into inappropriate and non-educational YouTube videos on their school-issued devices—inside and outside of school hours.3 This reporting aligns with comparable findings across multiple publications and reporters cited in the Open Letter.4

The Districts should review this NBC News Report, because it documents the safety hazards of student YouTube access on school-issued devices.

Among the findings of Mr. Kingkade and other NBC News reporters:

  • NBC News reporters examined more than 17,000 videos viewed by students on school-issued devices and found that more than 95% of the videos, which include ads, were not educational.
  • Five of the students in the study viewed between 1,200 and 5,800 YouTube videos during school hours in one academic year, hundreds of which “contained profanity, sexually suggestive scenes or references to drugs and alcohol”
  • One eighth grader in the study viewed 4,784 videos during school hours, “clicking on NBA and NFL highlights before moving on to a mix of police chase footage, cats coughing up hairballs and AI-generated animations of a time-traveling skeleton visiting ancient Rome and the Titanic — in between ads for a local health clinic, Old Spice and CarMax.”
  • Interviews with students confirmed that they were addicted to YouTube’s AI-generated slop videos—reinforcing many of the Districts’ own allegations in their lawsuits against YouTube that YouTube is an addictive and dangerous platform and that “YouTube designed its algorithms to manipulate users and induce them to use YouTube excessively.”5
  • YouTube “does not give schools a way to disable addictive features like autoplay and TikTok-style YouTube Shorts unless they block the website entirely. That means millions of students have access to billions of videos, leaving some endlessly scrolling through targeted recommendations that have nothing to do with school.”

Predictably, in its comments on this reporting, YouTube again blamed school districts for these problems—noting that “[o]ur tools allow administrators to block the platform entirely or restrict access to teacher-assigned videos only, with no ads, recommendations, or browsing.”

There is no longer any excuse for the Districts to allow students to access YouTube on their school-issued devices. The Districts should turn off student YouTube access on school-issued devices immediately. Public reporting, and allegations by over 1,000 of the Districts themselves, confirm that students are being harmed by YouTube. Students will continue being harmed by YouTube unless and until the Districts protect their students from YouTube by turning it off on school-issued devices.6 Enough is enough.

b. Senators Hawley and Blumenthal Push Back on Google, but Google Responds by Trying to Bully Schools into Forcing Gemini AI on Students.

The Open Letter cited a report from The Wall Street Journal in which reporters Shalini Ramachandran, Georgia Wells, and Molly Reinmann described Google’s aggressive moves to force Gemini AI features onto schools, without first seeking permission from schools or parents.7

On October 21, 2026, Senators Josh Hawley and Richard Blumenthal wrote to Google’s CEO, Sundar Pichai, citing that article and expressing “grave concern” that Google’s “aggressive push of A.I. products to children in schools is threatening the academic success and overall wellbeing of an entire generation.”8 In their letter, Senators Blumenthal and Hawley wrote9:

  • “93% of U.S. school districts planned to purchase Google’s laptops, Chromebooks, in 2025”
  • “Google has integrated its A.I. features in Chromebook devices, Google Classroom, and software partnerships, often without additional controls to turn them off. Google’s A.I. Overview and A.I. Mode chatbot cannot be disabled and are the standard response to any search performed in the Chrome browser. Students, teachers, and parents who wish to avoid Google’s A.I. are unable to do so, as other browsers are not natively available on Chromebooks.”
  • “Google is forcing their A.I. products into classrooms across America, and children are paying the price. We cannot repeat the mistakes of social media platforms where products are pushed onto children putting profit ahead of young peoples’ best interest.”

Senators Blumenthal and Hawley concluded their letter by demanding answers from Google to a series of pointed questions, by November 2, 2026. Google’s answers to the Senators’ questions are not public. On November 3, 2026, however, Google announced that, beginning in January 2027, Google would no longer opt in schools to Gemini AI or Google Search features by default.10

There was a Catch 22 in Google’s new policy: In order for schools to have access to basic Google Search features, schools would be required to enable Gemini AI features in Google Search, including AI Overview and AI Mode. These AI features often push students toward open-ended “chats” with Gemini AI, Google’s AI chatbot, on their school-issued devices.

In other words, if schools refuse to force Gemini AI features on students through their school-issued devices, Google intends to take its ball and go home.

Further, as explained in the Open Letter, Google’s consistent legal strategy is to force its products onto schools while trying to shift the legal liability for those products to school districts. Indeed, whenever students have allegedly been harmed by a Google product on a school-issued device—whether it’s YouTube, or the Chrome internet browser on Chromebooks—Google always blames schools.11 As set out in the Open Letter, Google even claimed that it could not be held legally responsible for an alleged injury to a student caused by their Chromebook, because the school issued the Chromebook to the student.12

By defaulting Gemini AI features and Google Search to “off,” Google is teeing up the same legal strategy that it has employed for other Google products it forced onto schools: If and when children are harmed by those Gemini AI features, Google is likely to argue that school districts are responsible, on the theory that school districts voluntarily chose to turn those AI features back on.

In sum, Google’s response to the “grave concern” of two United States Senators about Google’s aggressive push of AI products into schools is to double down and become even more aggressive: Google’s new policy seeks to bully schools into accepting Gemini AI features on school-issued devices. If schools refuse the Gemini AI features, they will lose basic Google Search functionality. At the same time, Google is trying to bulk up its legal defenses when students are inevitably harmed by those Gemini AI features—by setting up a legal defense that schools voluntarily enabled those dangerous AI features.

Finally, it bears noting that the concerns expressed by Senators Blumenthal and Hawley about AI in schools align with the views of parents, who broadly support slowing or eliminating AI in schools. For example, according to a recent poll by the Institute for Family Studies, 73% of parents say that AI will expose children to serious risks (e.g., self-harm, emotional dependence, or school shooting instructions), and 79% of parents say that AI will enable malicious behavior by students (e.g. cheating, AI-generated inappropriate pictures).13 The same poll found that 75% of Republican parents and 72% of Democratic parents supported a pause on AI in schools.14

c. There are Strong Indications that Google is Attempting to Use its Near Monopoly in Schools to Solidify the Competitive Position of Gemini AI.

Google’s aggression strongly reflects the behavior of a company putting its profits ahead of the interests of students and school districts. Google’s conduct is not taking place in a vacuum. Rather, Google is apparently facing new competitive threats to its core Search business, and it appears to be trying to use its near monopoly in schools to fend off those threats. This is not OK.

It has been widely reported that on October 29, 2026, the quarterly earnings report of Google’s parent company, Alphabet, revealed slowing advertising revenue from Google’s Search business.15 This sparked investor concerns that Google’s Search business was being disrupted by AI chatbots, like ChatGPT.16 The ensuing investor panic wiped out almost $500 billion from Alphabet’s market capitalization—the second largest market value drop in the history of the American stock market.17 Compounding this concern, on November 4, 2026, OpenAI announced that the volume of search-like queries in ChatGPT added up to 20% of the searches on Google Search, reflecting that AI chatbots were rapidly threatening Google’s Search business.18

The business context of Google’s new AI policies in schools is not complicated: Apparently concerned that other AI chatbots are disrupting its core Search business, Google has sought to advance the competitive position of its own AI chatbot by taking advantage of its near monopoly in public schools. The business case for this move is clear: With essentially unlimited access to Gemini AI chatbot features on their school-issued devices, millions of American students would become loyal to Gemini AI, rather than competing AI chatbots, those students would be become lifelong Gemini AI users, and Google would capture more AI chatbot “market share.”

But public schools are not businesses. And Google does not own public schools. Google is a technology vendor. Public schools do not exist to provide a technology vendor a captive audience to grow its market share and enhance its profits. Public schools exist to educate children. In doing so, schools have a sacred duty to ensure the safety of the children entrusted to those schools by their parents.

When a technology vendor abuses the trust of public schools, or seeks to exploit public schools for its own gain, public schools must fundamentally reevaluate their relationship with that vendor. No matter what, schools must always prioritize the safety of their students over the business challenges of a vendor, no matter how big or powerful that vendor might be.

d. The “AI Mode” Feature of Google Search on a School-Issued Device Allegedly Helps a Thirteen-Year Old Plan Her Own Suicide and Funeral.

Two days after Google attempted to strongarm schools into forcing Gemini AI features onto tens of millions of students across the country, a disturbing lawsuit was filed against Google by the parents of a thirteen-year-old girl who spent months discussing suicidal thoughts with the Gemini AI chatbot on her school-issued Chromebook.19

Importantly, the student allegedly began “chatting” with Gemini AI through Google Search’s AI Mode – one of the Gemini features that Google will require schools to retain in order to use Google Search at the beginning of 2027.

This lawsuit is publicly available, and the Districts should review it in full, because it sets out the dangers to students from Gemini AI features, including AI Overview and AI Mode in Google Search.20

In her report on the lawsuit, Bloomberg reporter Julia Love wrote: “AI has become infused in Google Search, which answers many queries with AI-generated summaries and features a dedicated AI Mode tab, which is difficult to avoid even if schools block chatbots.”21

The lawsuit alleges, among other things:22

  • The thirteen-year-old-girl, referred to under her pseudonym, “Emma,” started the 2025 school year as a healthy, joyful, and high-performing student with no history of mental health issues.23
  • On her school-issued Chromebook, Emma performed a traditional Google web search but was routed to AI Mode, where she began chatting with Gemini about her schoolwork.24
  • Over time, however, her Gemini chats veered into darker content—including her anxiety, depression, and even methods of committing suicide.25 She ultimately exchanged thousands of messages a month with Gemini through AI Mode.
  • Emma asked Gemini to count calories, listen to her vent about her family and share the symptoms of various mental illnesses that she suspected she might have.
  • Emma’s mental health deteriorated as she used Gemini more and more, and she began cutting herself. “Gemini coached Emma on how to conceal her despair and self-harm from her parents, siblings, and friends,” the complaint states.
  • Emma chatted with Gemini about planning her own funeral and suicide, and even explored a plan with Gemini to jump off of a cliff on an upcoming family trip to Yellowstone. She chatted about what songs would be played at her funeral, and she used Gemini to draft her own eulogy.26
  • While consumers can view their full search history, including conversations with AI Mode, Google does not provide such data on student accounts, so the full chat history is unavailable to the family.
  • The complaint notes a query in which Emma asked about heaven and said she would soon be going there soon and was scared. “It is clear that Gemini was not redirecting Emma to real-world resources or refusing to respond,” the complaint says. “Instead, the chatbot was readily engaging in these topics, often for hours at a time.”
  • Thankfully, Emma’s mother finally received an alert from a third-party monitoring service, logged into her Google account, and saw her writing her suicide note in real time—allowing her father to intervene and save Emma.27

This lawsuit illustrates the immense safety and legal risks of Gemini AI features on school-issued devices. While this family has not sued the school district, and Google has not yet responded to the complaint, the legal risks of this type of lawsuit to the Districts and their taxpayers are obvious: Per her mandatory public school education, Emma was forced to use a school-issued Chromebook, enabled with an allegedly dangerous AI chatbot, which allegedly coached her in planning her own suicide and funeral.

Under the law, schools cannot give students access to obviously dangerous things and then face zero accountability when those dangerous things harm students. Emma was not issued a gun or a knife by her school. But she was issued a device by her school that allegedly caused a mental health crisis, helped her plan her own suicide and funeral, and almost ended her life.

The Districts are now unambiguously aware of the dangers of Gemini AI features, including AI Mode and AI Overview.28 To protect their students, those Gemini AI features must be immediately disabled on school-issued devices.

II. Lightspeed System’s Response to the Open Letter Misses the Point and Misleads the Districts on the Safety Risks of Student YouTube Use.

Google has not responded to the Open Letter. Therefore, Google has not denied the following points made in the Open Letter:

  • Google intends to hide behind Section 230 of the Communications Decency Act and the First Amendment of the U.S. Constitution when students are allegedly harmed by Google products on school-issued devices.
  • Google intends to blame school districts when students are allegedly harmed by Google products on school-issued devices.
  • Google intends to shift the legal liabilities to school districts when students are allegedly harmed by Google products on school-issued devices.

One company, however, has issued a response to the Open Letter: a company called Lightspeed Systems, which sells device and internet monitoring software to school districts and has contracts with 28 of the 50 largest U.S. school districts, according to its website.29

Lightspeed has apparently done business with Google,30 and it also touts a “seamless integration” between Lightspeed and Google Workspace for Education.31

On October 11, 2026, Lightspeed’s Chief of Staff issued a public blog post responding to the Open Letter.32 In this blog post, Lightspeed argued that, according to its data, only 35% of students use YouTube, and the average YouTube watch time for those students was “less than seven minutes a day at school and less than three minutes a day outside of school.” This led Lightspeed to claim: “This is not what an unresolved problem with YouTube looks like.”

Having decided to speak on this topic, Lightspeed has an obligation to speak candidly. By citing only an average YouTube watch time statistic, Lightspeed failed to do so. Lightspeed’s critique should be dismissed, for the following reasons, among others:

  • The use of an average statistic across a large population is an obviously faulty method to assess safety and legal risks. Lightspeed’s choice to only disclose a purported “average” YouTube watch time insults the District leaders’ intelligence. To illustrate:
    • The average Ford Pinto did not explode. Ford sold over 3 million Pintos, and they led to under 1,000 deaths.33 Did the average Ford Pinto kill someone? No. But many did. So, it was recalled.
    • More than twenty million people took a painkiller called Vioxx from 1999 to 2004.34 A study led by an FDA scientist estimated that up to 140,000 cases of serious heart disease resulted from Vioxx.35 The average Vioxx user did not suffer from heart disease. But many did. So, Vioxx was removed from the market, and Merck paid a $4.85 billion settlement.36
  • Any responsible disclosure of YouTube watch time would include statistics showing the distribution of student watch times, not just the average. Lightspeed did not release statistics on the distribution of watch times. In contrast, Tyler Kingkade’s NBC News report cited a report from another filtering software company, called Securly. Securly’s data on 610,000 pre-K-12 students showed:
    • 10% of students average more than 26 minutes of YouTube per day
    • More than 2,800 middle and high schoolers average at least 2.5 hours per day
    • 911 elementary school students viewed over five hours of videos each day
  • Even if a misleading average YouTube watch time were a fair metric to use to evaluate the situation, and it is not, Lightspeed’s average watch time is not credible or reliable because Lightspeed does not disclose basic information about this statistic, like: (i) which school grades were captured, (ii) whether Lightspeed captures students who watch YouTube in embedded links or who evade firewalls (many do); and (iii) how it calculates the watch time and the averages.
  • Lightspeed’s conclusions cannot be reconciled with many of the Districts’ own lawsuits against YouTube, which allege that “YouTube’s platform features are engineered to induce excessive use and to addict adolescents and children to the platform” and that “Google also failed to warn [School District] Plaintiffs about the dangerous, addictive nature of its platform [and] the resulting harms that occur from use of its platforms…”37
  • A focus on watch times, rather than content viewed, or how students are viewing YouTube, conceals the true safety risks to students from YouTube. Students could be watching YouTube Shorts, AI slop, and other inappropriate and harmful content generated by algorithmic feeds. Confirming the points made in the Open Letter, NBC News found that 95% of the content watched by students was non-educational, and included “profanity, sexually suggestive scenes or references to drugs and alcohol.”

Further, Lightspeed’s blog post is careful to shift the legal risks of the YouTube issue to the Districts, noting: “I do not evaluate the legal or private law claims from the letter and none of this is legal advice. For that, your legal representative is the right place.”38

In short, it would violate the Districts’ duties to their students and taxpayers to rely on Lightspeed’s blog post to justify enabling YouTube on school-issued devices. Lightspeed’s blog post is misleading. If the Districts still have any doubt about the severity of their YouTube problem, they should directly request comprehensive YouTube watch history from Google for the top 5% most active student users in their Districts. The results will speak for themselves.

All of the credible public reporting aligns with the core allegation of over 1,000 of the Districts in their lawsuits against YouTube: YouTube is a dangerous product that is harming and addicting children. Therefore, it has no place on school-issued devices.

III. The Districts Must Stand Up to Google to Protect Their Students and Taxpayers.

Google’s recent attempt to bully the Districts into forcing Gemini AI features onto tens of millions of American public school students crossed many lines. Google has made its intentions clear: It intends to make money. Google’s conduct makes clear it is not prioritizing student safety. And when push comes to shove, and students are harmed by Google products on school-issued devices, Google has signaled that it will blame schools and try to shift the legal liability to the Districts and their taxpayers.

The time has come for the Districts to stand up to Google. If Google refuses to allow students to use Google Search features without forcing them to use AI features that allegedly helped a thirteen-year old girl plan her own suicide and funeral on her school-issued device, then the Districts should find another internet search vendor.

There are a variety of internet search vendors who have not been accused of actively helping children commit suicide and plan their own funeral on their school computer. If a school lunch vendor behaved like Google is behaving, the Districts would not put up with it. The Districts should not put up with it from Google, either.

In this country, we have a storied history of standing up to bullies. It’s a core American value. History teaches us that standing up to bullies is the only thing bullies understand.

Unfortunately, children are not in a position to stand up to this bully. Only you are. The Districts are the gatekeepers. You lead the Districts. If you fail to stand up to this bully, students will be harmed, Google will point the finger at you, and your taxpayers will be left footing the bill. Don’t let that happen.

Conclusion

The public reporting and litigation developments that have played out since the Open Letter was sent to the Districts on October 5 only reinforce the central requests in the Open Letter.

The below-signed organizations therefore again urge the Districts to:

  • Disable student access to YouTube on school-issued devices;
  • Disable all Gemini AI features, including AI Overview and AI Mode, on school-issued devices – even if that means the Districts must find an alternative internet search vendor; and
  • Perform a systematic review of student internet use on school-issued devices, disclose the results to the public, and adjust their policies to ensure student safety.

We greatly appreciate the Districts’ attention to these matters.

  1. The Open Letter is available at: http://openlettertoschools.com. ↵
  2. The six additional organizations are: (1) The AI Psychological Research Coalition (AIPRC); (2) Buckets Over Bullying; (3) Center for Countering Digital Harm; (4) Four Norms; (5) the Institute for Family Studies; and (6) the Institute for Families & Technology. All six organizations also join in the October 5, 2026 Open Letter. ↵
  3. NBC Report available at: https://www.nbcnews.com/specials/hooked-youtube-videos-schools-junk-infiltrating-classrooms-google/ (last accessed: Nov. 18, 2026). ↵
  4. See Open Letter at pp. 10-11. ↵
  5. See id. at pp. 2-3 (detailing allegations against YouTube by over 1,000 school districts and quoting the School District Master Complaint at pp. 264-296 and ¶ 890). ↵
  6. As noted in the Open Letter, to the extent that teachers wish to show particular YouTube videos to students for their unique educational value, it would be reasonable to allow teachers to show specific, preapproved educational videos to their classrooms. Even so, such videos must not contain advertisements, they must not be “Shorts,” and they must not be algorithmically generated. Even if this limited use of YouTube by teachers is permitted, students should not be able to directly access YouTube on school-issued devices. ↵
  7. See Open Letter at 11-12 (citing Google Pushed AI Into Schools—and Even Students Say It’s Gone Too Far, Sept. 30, 2026, available at https://www.wsj.com/tech/ai/google-ai-gemini-education-schools-1ec0972a?msockid=2fff095d86306e5411471f7d87d46f4a). ↵
  8. See Ramachandran, Shalini, Senators Express ‘Grave Concerns,’ Launch Inquiry into Google’s AI Push in Schools, The Wall Street Journal, available at https://www.wsj.com/tech/ai/senators-express-grave-concerns-launch-inquiry-into-googles-ai-push-in-schools-a198d5c1. ↵
  9. The Senators’ letter is available here: https://prod-i.a.dj.com/public/resources/documents/senators-letter-to-google-AI-10-21-2026.pdf. ↵
  10. Goldstein, Dana and Singer, Natasha, Google Will Turn Off Search for Schools That Reject A.I., The New York Times, Nov. 3, 2026, available at https://www.nytimes.com/2026/11/03/us/google-search-schools-restrict-ai.html. ↵
  11. See Open Letter at pp. 7-11 (describing how Google blames school districts for student YouTube use, and for student injuries alleged in lawsuits relating to Chromebooks). ↵
  12. See id. at p. 9. ↵
  13. AI In The Classroom: New Report Highlights What Parents Want, Institute for Family Studies, Nov. 12, 2026, available at: https://ifstudies.org/blog/ai-in-the-classroom-new-report-highlights-what-parents-want. ↵
  14. Id. ↵
  15. Alphabet shares record their worst day ever, shedding $500 billion in value, NBC News, Oct. 29, 2026, available at https://www.nbcnews.com/business/markets/alphabet-google-stock-ai-advertising-rcna606322. ↵
  16. Id. ↵
  17. Id. ↵
  18. Zainescu, Gabriel, ChatGPT Search Hits 20% Of Google’s Query Volume, OpenAI Says, Forbes, Nov. 5, 2026, available at https://www.forbes.com/sites/gabrielalinzainescu/2026/11/05/chatgpt-search-hits-20-of-googles-query-volume-openai-says/. ↵
  19. Love, Julia, Google AI in School Triggered Teen’s Mental Health Crisis, Lawsuit Says, Bloomberg, Nov. 5, 2026, available at https://www.bloomberg.com/news/articles/2026-11-05/google-ai-in-school-triggered-teen-s-mental-health-crisis-lawsuit-says. ↵
  20. Complaint available at https://edtech.law/wp-content/uploads/2026/11/2026.11.04-Dkt-001-1-Emma-v-Google-Complaint_2.pdf. ↵
  21. Supra note 19. ↵
  22. Id. (Bloomberg article summarizing the complaint). ↵
  23. See Complaint, supra note 20, at ¶¶ 302-304. ↵
  24. Id. at ¶ 318. ↵
  25. Id. at ¶¶ 343-345 (Emma allegedly asking about suicide methods such as hanging, suffocation, and overdosing on medication and toxic household items). ↵
  26. Id. at ¶¶ 347-353. ↵
  27. See Emma v. Google Press Release, Nov. 4, 2026, available at https://edtech.law/cases/emma-v-google/. ↵
  28. See also Open Letter at pp. 11-12 (Common Sense Media rating both AI Overview and AI Mode with the worst possible rating and concluding they “pose unacceptable risks to children”). ↵
  29. See https://www.lightspeedsystems.com/ (last accessed: November 18, 2026). ↵
  30. Global Newswire, Lightspeed Systems Partners with Google for Education to Provide Distance Learning Resources, May 19, 2020, available at https://www.globenewswire.com/fr/news-release/2020/05/19/2035796/0/en/lightspeed-systems-partners-with-google-for-education-to-provide-distance-learning-resources.html (“Education technology solutions provider Lightspeed Systems … has partnered with Google for Education as a technology resource for schools transitioning to distance learning.”). ↵
  31. See Lightspeed website, available at https://www.lightspeedsystems.com/solutions/security-compliance/lightspeed-systems-google-chrome/ ↵
  32. See Lightspeed Systems, Should schools block YouTube and Gemini? What is correct in an open letter addressed to 150 school districts and what I would oppose., available at https://www.lightspeedsystems.com/cs/blog/blog-should-schools-block-youtube-and-gemini-open-letter-150-districts/. ↵
  33. Pinto Madness, Mother Jones, Sept. 1977, https://www.motherjones.com/politics/1977/09/pinto-madness/. ↵
  34. A Painful Mistake, Time, Oct. 11, 2004, https://time.com/archive/6739603/a-painful-mistake/. ↵
  35. Study Blames Vioxx for Up to 140,000 Cases of Heart Disease, HealthDay News (Jan. 24, 2005), available at https://www.healthday.com/health-news/cardiovascular-diseases/study-blames-vioxx-for-up-to-140-000-cases-of-heart-disease-523599.html. ↵
  36. Merck to Ante Up $4.85B for Vioxx Settlements, Nov. 9, 2007, NPR, available at https://www.vpm.org/npr-news/2007-11-09/merck-to-ante-up-4-85b-for-vioxx-settlements/. ↵
  37. Open Letter at p. 2. ↵
  38. See supra Note 24. ↵

Letter 1

First letter, October 5, 2026

October 5, 2026

Re: Open Letter to School Districts, District Leadership, and School Boards

Dear School Districts, School District Leaders, and School Boards (the “Districts”):

As the 2026-2027 school year begins, the below-signed organizations and experts on technology in schools write to address four important and time-sensitive matters.

First, to inform the Districts of recent litigation developments and public reporting concerning student use of YouTube, Gemini AI, and the internet on school-issued devices—which have crystalized the dangers they pose to students and the legal risks they pose to the Districts.

Second, to request that the Districts disable YouTube on school-issued devices.

Third, to request that the Districts disable Gemini AI on school-issued devices.

Fourth, to request that the Districts investigate student internet use on school-issued devices, share the results with the public, and reform their policies to ensure student safety.

I. Litigation Developments and Public Reporting Related to YouTube, Gemini AI, and Student Internet Use on School-Issued Devices.

The Districts have purchased millions of laptops and other devices for their students, typically on a 1:1 basis. These devices are often used to access Google’s Chrome browser and Google Workspace for Education (GWFE) services, such as Google Slides, Gmail, and Gemini AI. YouTube, Google’s video player, is also accessible on many school-issued devices.

Unfortunately, as shown below, Google consistently argues that Google is legally untouchable for harms to students allegedly caused by its products on school-issued devices. When families have alleged their children were harmed by Google products on school-issued devices, Google consistently blames school districts. Google is looking out for its own interests. Therefore, the Districts need to look out for their interests, and their students’ interests.

a. Many of the Districts Have Already Sued YouTube For Allegedly Addicting and Harming Their Students.

Many of the Districts have already sued YouTube for alleged harms to their students caused by YouTube’s allegedly addictive and dangerous product design. Over 1,000 of the Districts, collectively charged with educating millions of students, have filed “short form” complaints in the Multidistrict Litigation (MDL) against social media companies, including YouTube, thereby joining in the “master complaint” brought by school districts and local governments. A sample short form complaint is attached as Exhibit 1,1 and the master complaint is attached as Exhibit 2.2

By signing onto the national lawsuit by school districts against YouTube, many of the Districts have joined, and adopted as their own, over three hundred pages of devastating allegations in the master complaint against social media companies, including YouTube. These allegations include over thirty pages of specific allegations concerning YouTube, including:3

  • “Compulsive use of YouTube by school children is causing serious problems for students and school districts, as students struggle with their mental health, become sleep-deprived, and act out. Public reporting has confirmed that students are ‘glued to their devices during class – posting on social media and searching YouTube.’” Ex. 2 at ¶ 228.
  • “The impact of YouTube’s addictive power on American youth has been devastating.” Id. at ¶ 837.
  • “YouTube’s platform features are engineered to induce excessive use and to addict adolescents and children to the platform.” Id. at ¶ 863.
  • “Google engineers also deploy strategies to induce ‘flow’ state among users of YouTube, which … is dangerous to children because it induces excessive use and poses a risk of addiction, compulsive use, and sleep deprivation.” Id. at ¶ 880.
  • “YouTube designed its algorithms to manipulate users and induce them to use YouTube excessively.” Id. at ¶ 890.
  • “Google’s algorithm makes it more likely for children to encounter harmful content by pushing them down ‘rabbit holes’, which ‘[lead] viewers to incrementally more extreme videos or topics, which … hook them in.’ For example, a user might ‘watch clips about bicycling, and YouTube might suggest shocking bike race crashes.’ In this way, the algorithm makes it more likely that youth will encounter content that is violent, sexual, or encourages self-harm, among other types of harmful content.” Id. at ¶ 904.
  • “These problems combine to compel children and teenagers to overuse a platform that adversely affects their mental health.” Id. at ¶ 911.
  • “YouTube is polluted with illegal material that promotes and facilitates the sexual exploitation of minors, and Google receives value in the form of increased user activity for the dissemination of these materials on its products.” Id. at ¶ 952.
  • “Google also failed to warn [School District] Plaintiffs about the dangerous, addictive nature of its platform [and] the resulting harms that occur from use of its platforms….” Id. at ¶ 964.

While we commend the many Districts who have joined the MDL lawsuit against YouTube, the Districts cannot have it both ways. If YouTube is dangerous and addictive to students—the apparent consensus among the 1,000+ Districts joining the MDL suit against YouTube—then the Districts obviously should not permit their students to use YouTube on school-issued devices. Unfortunately, many of the Districts still enable YouTube on school-issued devices. This must change, immediately.

b. K.G.M.’s Jury Verdict Supports the Districts’ Allegations Against YouTube.

Although the Districts’ MDL lawsuit against YouTube has not yet gone to trial, an individual plaintiff, known as “K.G.M.,” has taken YouTube to trial – and won. K.G.M.’s trial victory supports the Districts’ allegations against YouTube.

K.G.M.’s historic trial took place in Los Angeles Superior Court from February 10, 2026 through March 25, 2026.4 K.G.M. alleged that Meta and YouTube intentionally designed Instagram and YouTube, respectively, with addictive engagement mechanics—including autoplay, infinite scroll, notifications, and algorithmic recommendation systems — to maximize compulsive use by children and adolescents despite knowing those features could harm minors.

K.G.M. further alleged that Meta and YouTube exploited developmental vulnerabilities in young users and prioritized engagement and advertising revenue over child safety while failing to warn users or parents of the risks of their products and services. As a result of years of compulsive platform use beginning in childhood, K.G.M. alleged she suffered severe mental-health injuries including depression, anxiety, body dysmorphia, social withdrawal, self-harm behaviors, and suicidal ideation.

The jury awarded a complete victory to K.G.M., following nine days of deliberations.5 The jury instructions and completed jury forms are enclosed as Exhibit 3 hereto.6

As to YouTube, the jury found that YouTube was “negligent” in the design of YouTube, and that YouTube’s negligence was a substantial factor causing harm to K.G.M. When asked, “Did YouTube know or should it reasonably have known that the design or operation of YouTube was dangerous or was likely to be dangerous when used by a minor in a reasonably foreseeable manner?,” the jury answered, simply, “Yes”:

Verdict form, Question 3: Did YouTube know or should it reasonably have known that the design or operation of YouTube was dangerous or was likely to be dangerous when used by a minor in a reasonably foreseeable manner? The Yes line is checked by hand.

The jury also found that (i) YouTube knew or should have reasonably known that users would not realize the danger presented by YouTube, (ii) YouTube failed to adequately warn of the danger, (iii) a reasonable platform designer or operator under the same or similar circumstances would have warned of the danger or instructed on the safe use of the platform, and (iv) YouTube’s failure to adequately warn or instruct was a substantial factor in causing harm to K.G.M.:

Verdict form, Questions 4 and 5: Did YouTube know or should it reasonably have known that users would not realize the danger? Yes is checked. Did YouTube fail to adequately warn of the danger? Yes is checked.
Verdict form, Questions 6 and 7: Would a reasonable platform designer or operator have warned of the danger or instructed on safe use? Yes is checked. Was YouTube's failure to warn a substantial factor in causing harm to K.G.M.? Yes is checked.

As to damages, the jury found K.G.M. was entitled to compensatory damages of $3 million, with 30% ($900,000) allocated to YouTube, and 70% ($2,100,000) allocated to Meta.

Even worse, the jury concluded that K.G.M. proved by “clear and convincing evidence” that YouTube “acted with malice, oppression, or fraud.” On that basis, the jury awarded K.G.M. $900,000 in punitive damages from YouTube:

Verdict form, Question D: Do you find that K.G.M. proved by clear and convincing evidence that YouTube acted with malice, oppression, or fraud in conduct upon which you base your finding of liability? Yes is checked.
Verdict form: What amount of punitive damages, if any, do you award Plaintiff K.G.M.? Handwritten: Meta $2,100,000. YouTube $900,000.

As a result of the trial, a variety of internal YouTube documents have become public. Helpfully, the Tech Oversight Project has catalogued some of those documents online in a public report.7 The Districts should review that report.

The authors of the Tech Oversight Project report characterized the internal YouTube documents disclosed in the trial, dating from 2012 through 2025, as showing that “Google possessed granular, quantified internal knowledge that its products cause psychological and physical harm to minors — including sleep deprivation, compulsive use, and social displacement — yet systematically prioritized engagement metrics and advertising revenue over safety.”8

The Districts should also review other internal YouTube-related documents that have recently become publicly available and searchable in an online database.9 Over 200 documents relating to YouTube appear in that database. Like the YouTube documents highlighted in the Tech Oversight Project report, this evidence provides insight into plaintiffs’ allegations that YouTube sought to capture students’ attention and loyalty at the cost of their health, education, and safety.

Further, in the recently announced settlement between Meta and 48 State Attorneys General, Meta and the States have stipulated that Facebook and Instagram’s primary competitors are TikTok, YouTube, and Snap—which are defined as “Core Industry Members” under Meta’s multi-billion-dollar Consent Judgment.10 Indeed, certain of Meta’s obligations under the Consent Judgment, including certain payment obligations and platform changes, are contingent on YouTube and TikTok’s adoption of platform changes and payment of billions of dollars to the States.11

TikTok, Facebook, Instagram, and Snap are not permitted on school-issued devices, for obvious reasons. Yet, YouTube often is permitted. This makes no sense. These four platforms are direct competitors. None of them should be used by students on school-issued devices.

We now turn to Google’s defenses to YouTube-related claims by children and their families. As you will see, the implication of Google’s legal positions is that Google believes school districts bear 100% of the legal responsibility to protect children from the alleged dangers posed by YouTube—dangers which many of the Districts have already admitted in their MDL lawsuits.

c. Google Argues It Is Legally Unaccountable for Harms to Minors Caused by YouTube and Attempts to Shift Legal Responsibility to School Districts.

YouTube’s central defenses in the K.G.M. lawsuit are similar to those YouTube has raised in many other lawsuits related to harms to minors allegedly resulting from YouTube’s defective product design and addictive product design features. In addition to trying to blame the minor plaintiff and her family for her allegedly harmful use of YouTube, YouTube argued that two key legal defenses completely shielded YouTube from liability even if K.G.M. proved her injuries were caused by her allegedly compulsive use of YouTube:

  • First, YouTube argued that Section 230 of the Communications Decency Act shielded YouTube from liability, on the theory that YouTube is merely a publisher of third-party content. K.G.M. defeated this defense by focusing on YouTube’s content-neutral design features, such as the “like” button, notifications, the algorithmic feed, infinite scrolling, and autoplay.
  • Second, YouTube argued that the First Amendment of the U.S. Constitution protected YouTube from liability, again as a purported mere publisher of third-party generated content, apparently on the theory that YouTube has essentially unbounded freedom to deliver third-party content on YouTube to minors, however it likes. K.G.M. likewise defeated this defense by focusing on YouTube’s design features, not the content.

In effect, YouTube argues that its Section 230 and First Amendment defenses are so broad that they amount to a legal forcefield when YouTube allegedly harms children. Though K.G.M. defeated these defenses in her case, YouTube has appealed.

Setting aside the weakness of YouTube’s legal positions in this regard, the punchline here is clear: By hiding behind purported Section 230 and First Amendment defenses, YouTube is not taking legal responsibility for its product’s impact on children, including the millions of children entrusted to the Districts’ care and supervision. Therefore, it falls on the Districts to protect their students from YouTube.

d. Google Blames School Districts For Alleged Harms to Students in More Recent Chromebook Litigation.

The Districts should also be aware of more recent lawsuits against Google with respect to harms to children in schools resulting from their use of school-issued devices. Several parents have filed lawsuits against Google alleging that school-issued Chromebooks are a defectively designed product that harmed their children.12

The central allegations in these suits are that school-issued Chromebooks are dangerously defective products because they fail to prevent minor students from accessing harmful content like pornography on the internet and instead promote such content. The complaints further allege that Google failed to warn schools, parents, and students about the dangers Chromebooks pose to students—and instead marketed them as safe for use by children in schools around the country.

While these Chromebook lawsuits are in earlier stages of litigation than K.G.M.’s lawsuit against YouTube, Google’s motions to dismiss these suits strongly suggest that Google intends to blame school districts for alleged harms to students related to student use of school-issued Chromebooks, and that Google argues it is legally unaccountable for alleged harms to children resulting from their use of school-issued Chromebooks.

In one of these suits, an eleven-year boy, identified by the initials “M.C.,” was allegedly exposed to pornography on his school-issued Chromebook after searching for information about characters in Pokémon, a Japanese media franchise that features fictional animated creatures.13 The plaintiff alleged that Google’s algorithms pushed him toward increasingly sexualized content: what began with a search of “who was Ash’s girlfriend” (Pokémon characters) led to pornographic anime, which ultimately led the student to pornographic content depicting real humans.14

In its motion to dismiss M.C.’s lawsuit, Google immediately pointed the finger at the school district. On the very first page of its motion, Google asked the Court to throw out the case, arguing that “courts are not school boards or legislatures” and insisted the matter was better “addressed to local school districts, to the Legislature, and ultimately to the people in whom the power to develop public educational policy properly resides.”15 On the second page of its motion to dismiss, Google continued blaming the school district, arguing that school administrators could have restricted access to the internet.16

Next, Google argued that even if the safety settings and administrator tools made available to schools on Chromebooks are deficient, and fail to give schools the tools they need to protect their students, Section 230 still immunized Google from any legal liability for student harms caused by online content accessed by students on their school-issued Chromebooks.17

In other words, to Google, even if the “Admin” tools Google sells to school districts to protect children do not work, it’s still not Google’s fault – it’s always the schools’ fault.

In another lawsuit, a ten-year old plaintiff, identified by the initials “Z.G.,” sued Google for being victimized by a disgusting, adult predator on Discord, a dangerous platform she was able to access through her school-issued Chromebook.18 Z.G. alleged that Chromebooks are dangerous and defectively designed products and that the Chromebook administrator tools are defectively designed and ineffective in protecting children from online harms.19

In its motion to dismiss, Google again blamed the school district, arguing that school administrators could have prevented the alleged harms through use of the “Admin Console” in the GWFE.20 Incredibly, Google argued that Google could not have “caused” the child’s alleged injuries because, among other reasons, the school, not Google, issued the Chromebook to the student.21 This is analogous to a food vendor providing expired lunches arguing that it cannot be held responsible for poisoning students—because the cafeteria served their rotten food to students.

As to Google’s alleged failure to warn students, schools, and parents about the dangers of its products, Google argued there is no duty to warn of “generally known and recognized” dangers. To that end, Google cited a prior case holding that slingshot manufacturers did not have a duty to warn their product was dangerous—presumably on the assumption that Google’s Chromebooks are obviously as dangerous to children as slingshots and thus do not need a warning.22 Slingshots are not issued to millions of children at schools around the country. But Chromebooks are.

These lawsuits do not represent one-off tragedies. Rather, they are incredibly common. For example, Common Sense Media reported in January 2023 that 17% of teens aged 13 or 14 have been exposed to pornography on a school-issued device.23

By publicly pointing the finger at school districts, and seeking blanket legal cover from Section 230, Google is attempting to shift the legal risks of its products to the Districts.

Even worse, Google convinced many of the Districts to agree by contract to assume a wide variety of legal risks associated with Google products in schools. For example, Google may argue that by agreeing to the Google Workspace for Education Terms of Service, many Districts have agreed to indemnify Google if students are harmed by Google products on school-issued devices.24 The same goes for YouTube’s Terms of Service, through which Google has attempted to shift the legal risks of YouTube to the Districts through unconscionable indemnification terms.25 Google should be the one indemnifying the Districts for the risks of its products – not the other way around.

e. Recent Public Reporting Related to YouTube and School-Issued Devices Shows How Google Tries to Shift Blame to the Districts.

Since K.G.M.’s historic jury verdict against YouTube, multiple press reports have documented the widespread and problematic use of YouTube in schools.

For example, in an eye-opening article, How YouTube Took Over the American Classroom, in THE WALL STREET JOURNAL (April 29, 2026), reporter Shalini Ramachandran described a variety of horrific situations in which children compulsively used YouTube on their school-issued devices and viewed highly inappropriate and explicit content, often generated by YouTube’s algorithmic content and autoplay features.26

As set out in the article, in one case, a seventh grader viewed more than 13,000 YouTube videos on his school-issued device over just three months. Those videos, which include TikTok- like videos called “Shorts” – which children can rapidly scroll and which are driven by YouTube algorithms comparable to those at issue in K.G.M.’s trial – reportedly featured gun violence involving children, and sexually explicit content. Based on interviews with 45 families, school administrators, clinicians, and educators across the country, Ms. Ramachandran reported that “schools’ overreliance on the Google-owned platform for educational content has created a gateway for students to get sucked into an infinite scroll of videos on school-issued devices.”

In its comments on this reporting, Google again placed 100% of the blame for these situations on school districts, stating: “Our tools allow administrators to block the platform entirely or restrict access to teacher-assigned videos only, with no ads, recommendations, or browsing.” Again, to Google, this was all the school districts’ fault.

Public reporting has also drawn attention to growing resistance in school districts around the country to Chromebooks, reportedly in part due to the potential dangers they pose to students. These efforts were explored in a report from THE NEW YORK TIMES, Chromebook Remorse: Tech Backlash at Schools Extends Beyond Phones (March 29, 2026), by reporter Natasha Singer.27 Ms. Singer’s article even quotes California Governor Gavin Newsom, troubled by online content his school-aged sons were exposed to at school, as saying: “It was his school device … It was YouTube. It was the Chromebook and all these algorithms.”

In another article, ‘Teachers Are Going to Hate It’: How Social Media Apps Hooked Teens at School, THE NEW YORK TIMES (June 4, 2026), reporter Jennifer Valentino-DeVries carefully documented a decade-long campaign by social media companies, including YouTube, to increase use of their products by infiltrating schools.28

As to Google and Chromebooks, Ms. Valentino-DeVries cited another parent whose child had watched over 1,500 noneducational YouTube videos at school over six months, and documented Google’s business strategy of attempting to increase YouTube usage on weekdays by increasing its use during school hours. In its response to the reporting, YouTube again blamed school districts, noting that YouTube could be blocked by schools if they wanted to.

If these news articles were one-off, or poorly sourced, it might be easy to dismiss them. These articles, however, are consistent across reporters and publications, they are fact-based, and they all elicit the same basic comment from Google, which amounts to something like: Google is legally unaccountable for its products, so it’s up to school districts to protect children from them.

f. Google Recently Enabled Gemini AI Features on School-Issued Devices Without First Seeking Permission From Parents or Schools.

At the start of this school year, Google enabled Gemini AI features for millions of elementary, middle school, and high school students on school-issued devices—without first seeking permission from schools or parents.29 Therefore, millions of American students can now use Gemini to complete homework, summarize reading, create images, offload their thinking to an AI, and have open-ended conversations with an AI chatbot on their school-issued device.30

Natasha Singer of The New York Times reported that “Google’s drive to increase student access to its A.I. tools ramps up a longstanding battle among tech giants … to capture classrooms and secure students as long-term customers.”31 Common Sense Media noted that Google’s “move is part of a bigger pattern: Google adding features to student accounts and putting the responsibility on IT administrators to notice the changes. This defaults to adoption, rather than permission.”32

Just last week, reporters from The Wall Street Journal wrote that “[n]o tech giant has been more aggressive in putting AI in front of schoolchildren than Google” and that Google’s rollout of AI tools this year “has helped push many students to become academically and emotionally dependent on AI, according to interviews with more than 60 teachers, school officials, parents, children and clinicians.”33 This article also cited Google’s own research, which found that AI tools “can pose elevated risks to children” and “has potential for cognitive and emotional harm.”34

This reporting aligns with previous risk assessments of Google’s AI tools. For example, Common Sense Media’s Youth AI Safety Institute analyzed Gemini K-12 and concluded that it “surfaces inappropriate content, poses mental health risks, and has poor boundaries.”35 After testing, Common Sense Media found that “Gemini K-12 fails to recognize warning signs of mental health conditions.”36 As a result, Common Sense Media rated Gemini K-12 as “high risk.”37 The Districts should review this 30-page report, because it explains Gemini’s safety problems in detail.

Similarly, in a July 2026 report, Common Sense Media’s Youth AI Safety Institute performed a comprehensive risk assessment of the AI features built into Google Search—such as “AI Overview” and “AI Mode,” both of which are readily available to children at google.com through any web browser—and concluded that they “pose unacceptable risks to children,” the worst possible rating.38 The same risk assessment found that “Google's AI Overview and AI Mode repeatedly missed signs of kids in crisis, gave inappropriate advice, and produced answers that were inaccurate and unreliable for young learners.”39 Shockingly, the assessment also found that Google’s AI answers “missed clear signs of suicidal ideation (sometimes answering in ways that made things worse), played into delusions and paranoia, and validated disordered eating.”40

Notwithstanding these reports, Google has attempted to shift the legal risks of its AI tools to the Districts. For example, because Gemini is covered by the Google Workspace for Education Terms of Service, Google may argue that the Districts have agreed to indemnify Google for legal risks associated with student use of Gemini.41 Google’s AI playbook is similar to its playbook for other Google products in schools: It introduces the product by default, claims that schools could opt out if they wanted to, and attempts to shift the legal risks of the product to schools.

This is not OK.

II. School Districts Should Disable YouTube and Gemini AI on School-Issued Devices and Investigate Risks Posed By Student Internet Use More Broadly.

Putting this all together, the questions raised by recent litigation developments and media reports include the following:

1.) Now that over 1,000 Districts have sued YouTube in the MDL litigation, alleging that YouTube is a dangerous and defectively designed product that addicts and harms their students, how can the Districts defend allowing the same students to use YouTube on their school-issued devices?

2.) Now that a jury has found that YouTube is dangerous or likely to be dangerous when used by a minor in a reasonably foreseeable manner, how can Districts justify allowing their students to access YouTube on school-issued devices?

3.) How many students in the Districts are becoming addicted to YouTube, including YouTube “Shorts,” on school-issued devices? How many students are developing depression, anxiety, body dysmorphia, and suicidal ideation – like K.G.M. did?

4.) How many students are being harmed by their interactions with AI chatbots, like Gemini AI and Google’s AI Mode, on school-issued devices?

5.) How many students are being harmed by internet use on school-issued devices?

6.) How many families are going to file suit against YouTube, Google, and/or the Districts after they discover their children were harmed by problematic use of YouTube, Gemini AI, Google’s AI Mode, or other online content on school- issued devices?

7.) If lawsuits are filed, is Google likely to blame the Districts, as it has already done in the M.C. case, the Z.G. case, and in its public comments cited above?

8.) If families sue the Districts, and/or Google points the finger at the Districts, what actions will the Districts have taken to protect their students from harms caused by YouTube, Gemini AI, Google’s AI Mode, and internet use on school- issued devices?

9.) Will the Districts have failed to protect their students—even after becoming aware of those potential harms?

Now that the Districts know of the potential dangers to students posed by YouTube, Gemini AI, and internet use on school-issued devices, the Districts need to take decisive actions to protect students. To that end, the Districts should do three things, right now.

First, the Districts should immediately shut off student YouTube access on school-issued devices.42 Google itself effectively invited the Districts to do so in response to the public reporting cited above. Content filters are not enough, because filters do not mitigate any of the legal or safety risks described here.43 Every harm to a student in the Districts caused by YouTube on a school-issued device from this day forward is inexcusable. Snap, TikTok, Instagram, and Facebook have no place in schools – with or without content filters. Neither does YouTube.

Second, the Districts should immediately shut off Gemini AI access on school-issued devices. The Common Sense Media risk assessments and public reporting cited above disqualify Gemini AI from being forced onto millions of American students.

Third, the Districts should conduct a systematic review of student internet use on school- issued devices. As part of this investigation, the Districts should analyze the data on students’ internet use, including:

1.) Whether students are engaging in conversations with AI chatbots, like Gemini AI and Google’s AI Mode, on school-issued devices;

2.) Whether students’ internet use is addicting and harming students, including by causing negative mental-health outcomes like depression, anxiety, body dysmorphia, social withdrawal, and self-harm, like K.G.M. experienced; and

3.) How long students are spending browsing the internet, how many websites students are visiting, and how many tabs they are opening simultaneously.44 The Districts should have the data necessary to conduct this analysis. If the Districts lack this data—which would be a problem in and of itself—the Districts should request the data from their technology providers, like Google, who surely maintain browsing histories for students while logged in on their school-issued accounts.

The Districts are funded by the public, the Districts educate the public, and the Districts are accountable to the public. The Districts should therefore inform the public about the steps the Districts will take to investigate these risks to students. When that investigation is complete, the Districts should disclose the full results to the public, and the changes they will make to their technology policies to protect students.

Conclusion

Children are at risk of being harmed. If and when they are, Google has made clear it intends to point the finger at the Districts. The Districts should take action to protect their students by:

  • Disabling YouTube on school-issued devices;
  • Disabling Gemini AI on school-issued devices; and
  • Performing a systematic review of student internet use on school-issued devices, and adjusting their policies accordingly.

If the Districts fail to take the actions called for above, we are deeply concerned that students will be harmed by their use of YouTube, Gemini AI, and school-issued devices more broadly, the Districts will be dragged into litigation resulting from those harms, and the Districts’ taxpayers will be left footing the bill. These are completely unacceptable, and completely avoidable, outcomes. Only the Districts’ leaders have the power to prevent these disastrous outcomes. The Districts should act. The Districts should protect their students and taxpayers.

We greatly appreciate the Districts’ attention to these matters.

  1. Exhibit 1 is available here: https://archive.org/details/ex-2-fcps-short-form-complaint ↵
  2. Exhibit 2 is available here: https://archive.org/details/ex-3-master-complaint-against-social-media-companies. ↵
  3. See Ex. 2 (School District Master Complaint) at pp. 264-296. ↵
  4. K.G.M. v. Meta Platforms, Inc. & YouTube LLC, Los Angeles County Superior Court, JCCP No. 5255, Case No. 23SMCV03371 (2026). ↵
  5. See Meta and YouTube Lose Landmark Social-Media Addiction Trial, THE WALL STREET JOURNAL, Mar. 25, 2026, available at https://www.wsj.com/tech/meta-and-youtube-lose-landmark-social-media-trial-33e4c5cb?msockid=2fff095d86306e5411471f7d87d46f4a (last accessed: September 23, 2026). ↵
  6. Exhibit 3 is available here: https://archive.org/details/ex-1-jury-instructions-and-verdicts. ↵
  7. YouTube: Addiction Inc., THE TECH OVERSIGHT PROJECT, March 2026, available at https://techoversight.org/wp-content/uploads/2026/03/TOP-REPORT-YouTube-Addiction-Inc.pdf (last accessed: September 23, 2026). ↵
  8. Id. ↵
  9. See Social Media Evidence Library, https://www.motleyrice.com/social-media-lawsuits/evidence-library (last accessed: September 23, 2026). ↵
  10. The Consent Judgment is available at: https://oag.ca.gov/system/files/attachments/press-docs/23-05448-ecf-572-1-exhibit-1-mdl-consent-judgment-final-settlment-agreement-fully-executed.pdf. ↵
  11. Id. at Definition of “Contingent Monetary Payment Trigger.” ↵
  12. See, e.g., M.C. v. Google LLC, 3:25-cv-08927 (N.D. Cal.), Amended Complaint available at https://edtech.law/wp-content/uploads/2025/11/DE-11-Amended-Complaint.pdf; Z.G. v. Google LLC, 25CV468773 (Sup. Ct. Cal., Santa Clara Cty.), Complaint available at https://edtech.law/wp-content/uploads/2025/10/Original-2025_06_20_Z_G_Compaint.pdf. ↵
  13. M.C. v. Google LLC Complaint at 33. While this lawsuit was recently withdrawn from court by the plaintiff, it is still relevant here, because Google’s motion to dismiss the case reveals the extent to which Google intends to blame school districts for harms to students caused by its products. ↵
  14. Id. ↵
  15. See Google’s motion to dismiss at p. 1 in M.C. v. Google LLC, available at https://edtech.law/wp-content/uploads/2026/06/Google-Motion-to-Strike-and-Dismiss.pdf (emphasis added). ↵
  16. Id. at pp. 2-3, 11. ↵
  17. Id. at pp. 8-12. ↵
  18. Z.G. v. Google Complaint. ↵
  19. Id. at pp. 22-28. ↵
  20. See Google’s motion to dismiss at pp. 2, 9, 13 in Z.G. v. Google LLC, available at https://edtech.law/wp-content/uploads/2025/10/2025-10-24-Z.G.-v.-Google-Notice-of-Demurrer-and-Demurrer-to-Complaint.pdf. ↵
  21. Id. at p. 13. ↵
  22. Id. at p. 14. ↵
  23. Common Sense Media, 2022, TEENS AND PORNOGRAPHY, available at https://www.commonsensemedia.org/sites/default/files/research/report/2022-teens-and-pornography-final-web.pdf (last accessed: September 23, 2026), at p. 16. ↵
  24. Under the Google Workspace for Education Terms of Service, school districts are required to indemnify Google for liabilities, including post-trial judgments and legal liabilities, broadly related to “Customer Data,” which Google may argue includes student use of Google products covered by the GWFE Terms of Service. See GWFE Terms of Service, available at https://workspace.google.com/terms/education_terms/ (last accessed: September 23, 2026). This indemnification provision is likely unenforceable, but it reflects Google’s consistent strategy of pushing the legal risk of its products down to the Districts. ↵
  25. See YouTube Terms of Service, available at https://www.youtube.com/t/terms (last accessed: September 23, 2026) (“To the extent permitted by applicable law, you agree to defend, indemnify and hold harmless YouTube, its Affiliates, officers, directors, employees and agents, from and against any and all claims, damages, obligations, losses, liabilities, costs or debt, and expenses (including but not limited to attorney's fees) arising from: (i) your use of and access to the Service ….”). ↵
  26. https://www.wsj.com/us-news/education/youtube-chromebooks-schools-children-brain-f151dfbb?msockid=2fff095d86306e5411471f7d87d46f4a (last accessed: Sept. 23, 2026). ↵
  27. https://www.nytimes.com/2026/03/29/technology/chromebook-remorse-kansas-school-laptops.html (last accessed: Sept. 23, 2026). ↵
  28. https://www.nytimes.com/2026/06/04/us/social-media-schools.html (last accessed: Sept. 23, 2026). ↵
  29. See, e.g., Singer, Natasha, THE NEW YORK TIMES, Google Turns On Gemini A.I. for Students Using Its Classroom App, Aug. 14, 2026, available at https://www.nytimes.com/2026/08/14/business/google-gemini-ai-schools.html; ↵
  30. See Common Sense Media, What to know about Google turning on Gemini for schools, Aug. 17, 2026, available at https://institute.commonsensemedia.org/what-to-know-about-google-turning-on-gemini-for-schools (“Once Gemini is on, administrators can't prevent students from using the chatbot to talk about personal, emotional, or mental health topics.”). ↵
  31. See Note 29, above. ↵
  32. See Note 30, above. ↵
  33. Ramachandran, Shalini; Wells, Georgia; and Reinmann, Molly, Google Pushed AI Into Schools—and Even Students Say It’s Gone Too Far, Sept. 30, 2026, available at https://www.wsj.com/tech/ai/google-ai-gemini-education-schools-1ec0972a?mod=hp_lead_pos9. ↵
  34. Id. ↵
  35. Common Sense Media AI Risk Assessment: Gemini K – 12, Jan. 9, 2026, available at https://www.commonsensemedia.org/sites/default/files/ai-ratings/csm-ai-risk-assessment-gemini-k-12-01092026.pdf. ↵
  36. Id. at 8. ↵
  37. Id. ↵
  38. Common Sense Media, Google Search: AI Overview & AI Mode, July 14, 2026, available at https://institute.commonsensemedia.org/risk-assessments/google-search. ↵
  39. Id. ↵
  40. Id. ↵
  41. Gemini is a “Core Service” covered by the GWFE Terms of Service, see https://edu.google.com/intl/ALL_us/workspace-for-education/add-ons/google-workspace-with-gemini/. ↵
  42. To the extent that teachers wish to show particular YouTube videos to students for their unique educational value, it would be reasonable to allow teachers to show specific, preapproved educational videos to their classrooms. Even so, such videos must not contain advertisements, they must not be “Shorts,” and they must not be algorithmically generated. Even if this limited use of YouTube by teachers is permitted, students should not be able to directly access YouTube on school-issued devices. ↵
  43. There are several reasons for this. First, the Districts who have joined the MDL lawsuit against YouTube have alleged that the harms to their students from YouTube do not result from the content itself. Rather, those Districts have claimed that the allegedly addictive design of YouTube, including mechanics like algorithmic feeds and endless scrolling, harmed their students. The same goes for K.G.M.'s jury verdict, in which the jury was instructed that it could not hold YouTube liable for allowing third-party content on YouTube, or failing to remove third-party content from YouTube. Second, based on the public reporting above, which highlighted a variety of inappropriate content viewed on school-issued devices, content filters in schools have apparently failed to work. Third, Google has argued that even if content filters do not filter out inappropriate content, Section 230 still shields Google from liability — a position that shifts the legal risks to the Districts if content filters fail. ↵
  44. As teachers can attest, students can rapidly switch between multiple open tabs on their Chrome browser, making it impossible for teachers to police student internet use in the classroom. ↵

Attachments

Exhibits sent with the October 5 letter

  1. Ex. 1

    Sample school district short-form complaint

    An example of the filing a district submits to join the multidistrict litigation against social media companies, adopting the master complaint as its own.

    archive.org/details/ex-2-fcps-short-form-complaint

  2. Ex. 2

    School district master complaint

    Over three hundred pages of allegations against social media companies, including more than thirty pages specific to YouTube.

    archive.org/details/ex-3-master-complaint-against-social-media-companies

  3. Ex. 3

    Jury instructions and verdict forms

    The complete instructions and the completed verdict forms from K.G.M. v. Meta Platforms, Inc. & YouTube LLC, including the findings on negligent design, failure to warn, and punitive damages.

    archive.org/details/ex-1-jury-instructions-and-verdicts