CCDH Files Amicus Brief, Claiming Social Media Could Have Radicalized the Buffalo Shooter 

CCDH is getting involved in a legal battle over whether the algorithm can be put on trial 

Before he murdered ten people at a Buffalo supermarket, a young Payton Gendron complained in his diary about a problem plaguing so many teenagers—he couldn’t get off his phone. He wrote: 

“Why do I always have trouble putting my phone down at night? . . . I’m a literal addict to my phone. I can’t stop consuming.” 

The way Payton described his compulsive phone use appears to reflect the very harm many like him have been facing: social media algorithms programmed to do anything to keep young people scrolling for as long as possible.  

In his case, those consequences were deadly.  

In the 180-page manifesto he published online, Payton explained he chose his victims based on a racist conspiracy he absorbed online—the Great Replacement theory—which claims non-white immigrants are deliberately replacing white populations in Western countries.  

Additional logs from his online presence, including on Discord, revealed his increasingly violent and extremist thinking with vocal support to previous mass-shooters alongside proof of him planning the attack. 

Survivors and the victims’ families sued the social media platforms, claiming algorithms, infinite scroll, autoplay, and other engagement-maximizing design features helped feed that debunked theory to a susceptible young user, leading him deeper into an online world of racism and violence. 

This fall, New York’s highest court will decide whether that case gets heard by a jury at all after an appeals court threw it out, citing a get-out-of-jail-free card reserved for Big Tech called Section 230 of the federal Communications Decency Act. 

CCDH filed an amicus brief in the case, providing an expert perspective to the court on why this case needs to be heard and why Section 230 should not be interpreted as a broad legal shield for social media companies.  

Section 230 was written for a very different internet 

To understand why these claims could be dismissed before the families get the chance to prove them, you need to understand Section 230. 

In 1996, years before Facebook, YouTube, Instagram, or TikTok existed, Congress passed Section 230 with the dual intent of protecting minors from sexually explicit content and the owners of websites like messaging boards from being sued over individual users’ potentially unlawful posts.  Today’s social media platforms do far more than provide somewhere for users to post. 

These companies have advanced to operate powerful addiction algorithms that study every pause, click, reaction, and search to predict what will keep each person engaged and what content should come next.  

At its core, the law says an online service generally cannot be held liable as the “publisher” for content created by somebody else. As written, Section 230 says nothing about strict product liability, negligent design, or the duty to warn users. 

Over the past three decades, courts have struggled to interpret Section 230 with circuit splits across the country. At times, the courts go too far and treat Section 230 like an expansive shield insulating the world’s wealthiest companies from otherwise meritorious lawsuits by their victims. In other cases, the courts have found liability for product designs or in other instances. The families in Patterson argue that social media companies should not be able to invoke a dated law to avoid scrutiny of their own design decisions. 

How the algorithm and other addictive features feed radicalization  

Big Tech companies often tell us that algorithms simply respond to what users want, but CCDH’s research tells a different story. The shaping starts before you’ve knowingly shown the algorithm anything. 

In our Deadly by Design study, we registered new TikTok accounts as 13-year-old users. The “For You” feed served them suicide-related content within 2.6 minutes.  

Accounts whose only distinguishing feature was a username hinting at vulnerability —  “loseweight” for example — received twelve times as many self-harm videos as standard accounts. Based on only that one signal, the algorithm saw a vulnerability and preyed on it. 

That is the mechanism the Buffalo families describe and want accountability for: a system that finds a vulnerable user and builds them a pathway to extremism.  

Safety is a setting and companies control the dial 

These features do not appear out of thin air. People build and test them.  

Internal documents show platforms know exactly what their ranking systems reward. Facebook’s engagement formula at one point scored an “angry” reaction at five times the value of a “like.” Anger was turned into the product and monetized.  

Then there’s Project Daisy, an experiment on the algorithm by Meta. According to a complaint filed by 33 state attorneys general, Meta’s own 2020 study showed that hiding “Like” counts reduced negative social comparison among teenagers. Meta’s leadership declined to roll it out as a default, after estimating it would cost about 1% of advertising revenue. 

And when Meta announced in January 2025 that it would stop proactively enforcing rules against violence and incitement, hateful conduct, and bullying, we measured what happened next: abusive comments tripled. 

Safety is a setting. These companies test it, price it, and decide where to set the dial.  

When a safer product is less engaging or threatens advertising revenue, safety loses.  

We’re asking for a trial, not a verdict 

Our brief does not ask the court to hold any platform liable, remove any content, or weaken protection for anyone’s speech. It simply asks that claims about the platforms’ own design choices survive long enough for the families to investigate what happened. 

Supreme Court Justice Clarence Thomas has previously made a similar point about the sweeping legal immunity courts have given platforms under Section 230:  

“Paring back the sweeping immunity courts have read into §230 would not necessarily render defendants liable for online misconduct. It simply would give plaintiffs a chance to raise their claims in the first place.” 

Social media algorithms and other engagement-maximizing features can lead a vulnerable teenager from curiosity to obsession and from racist propaganda to real life violence.  

When a company is alleged to have engineered addiction into a product, targeted it at adolescents, and known that its escalation dynamics steered vulnerable users toward violent extremism, it cannot be said as a matter of law, on the pleadings, and with no record that violence by one of those users was “extraordinary” or “far removed” from the design. 

The Buffalo families deserve to know whether that is what happened here. 

Before Big Tech is given immunity, the public should ask what the platforms knew, what their systems did, and whether different choices could have interrupted the path to violence. 

Read CCDH’s full amicus brief in Patterson v. Meta— and sign up for updates on this legal battle and others.