In an age of growing divide between the Democratic and Republican parties, it is rare to see a bill co-authored and advocated for by multiple members of both sides. Yet who could predict that big tech companies would be the unifying force that caused members of both political parties, with the help of testimony from A-list Hollywood actors, to work together on a piece of legislation? 

In December 2025, Senators Lindsay Graham (R-SC) and Dick Durbin (D-IL) introduced a bill to reform Section 230 (§230) [1]. Section 230 of the Communications Decency Act has historically protected tech companies from lawsuits against malicious third-party content on their platforms that is excessively violent, harmful, or threatening to users [2].

Presently, because of harmful social media messaging and smartphone usage, roughly 2,000 additional teen suicides occur every year [3]. Because of the large death count from social media usage, some opponents argue that the lack of legal liability for unmoderated harmful content on big tech and social media platforms is unacceptable.

There are rising rates of anxiety and depression amongst youth. 46% of adolescents aged 13-17 say social media makes them feel worse [4]. To address these rising rates stemming from adolescents’ use of social media platforms such as TikTok and YouTube, the bill to “sunset” §230, which would make tech companies legally and financially liable for malicious content posted on their platforms, was introduced [5].

Critics have also pointed toward capture theory—the theory that the interests of tech leaders may be captured by political lobbyists, resulting in them taking down reasonable content that does not align with their political interests—as another reason why social media giants should be more strictly regulated [6].

The Computer & Communications Industry Association estimated that repealing §230 would “cost digital services and their users more than $100k in legal fees per case totaling to $110 billion per year” because a single lawsuit typically costs over $100,000 in discovery fees [7]. Presently, there is no minimum or maximum amount set for how much a tech corporation could be sued for if Sunset §230 passes.

While §230 must be reformed, a complete dismantling of §230 may destroy free speech across all media platforms, even those not belonging to tech giants, due to tech companies becoming increasingly averse to legal entanglements. I propose instead that the language of §230 be refined to remove immunity for algorithms and generative AI deployed by social media companies that amplify violent content.

What’s in Section 230?

Originally written to help tech companies grow without facing a deluge of lawsuits, §230 now faces controversy over its explicit immunity for big tech companies. There are two main provisions of Section 230 that make the statute particularly controversial. The first is from §230(c)(1), which specifies that service providers and users may not “be treated as the publisher or speaker of any information provided by another information content provider” [8]. In short, because tech platforms are not treated as the publisher or speaker of information, they are immune from civil lawsuits concerning controversial posts uploaded by their platform’s users.

Opponents of §230(c)(1) argue that without any explicit legal liability, tech providers have no pressure to remove hate speech and disinformation in a timely manner [9]. However, §230 has key exceptions under which tech companies could be held liable, including for federal criminal law violations, intellectual property claims (e.g., copyright), Electronic Communications Privacy Act (ECPA) violations, and sex trafficking. Section 230 also protects tech companies from lawsuits for voluntarily acting in good faith to make restrictions on “material that the provider or user considers to be obscene, lewd, lascivious, filthy, excessively violent, harassing, or otherwise objectionable” [10]. 

How Has Section 230 Been Applied in Courts?

Several legal contentions have applied §230 in their rulings. In one 2009 case, Goddard v. Google, the plaintiff sued Google for displaying third-party ads for fraudulent mobile subscription services [11]. The court dismissed the plaintiff's claims and held that §230 barred the suit because the search feature was a neutral tool that did not make the advertiser solely responsible. Additionally, in the 2023 case of Twitter v. Taamneh, the Supreme Court found that Twitter was not liable for aiding ISIS after it was accused of recommending ISIS content to users, following the Supreme Court’s tradition of judicial minimalism in cases adjacent to §230 [12]. These rulings are viewed as particularly controversial due to the role the content in question had in potentially perpetuating mass violence and fraud content posted on Google and Twitter.

The future of Section 230 will also be heavily impacted by the 2026 lawsuit brought forth by 1,600 plaintiffs, including families and school districts, against Meta and YouTube. In this case, decided by the Los Angeles Superior Court, a $6 million verdict was entered in favor of the plaintiffs against Meta and YouTube [13]. This was due to the addictive design of the platforms leading to the plaintiffs’ exposure to content relating to body image and self-harm, which the plaintiffs alleged worsened their depression and anxiety [14]. This landmark decision has the potential to fuel the fight to sunset §230, inspire more lawsuits, and weaken §230 immunities for tech companies.

Implications of Sunsetting §230

One major concern critics have about reform efforts to §230 is the risk of infringement on users’ freedom of speech on social media platforms. Tech companies may feel compelled to moderate everything, including lawful but controversial content, to avoid potential lawsuits. Yelp’s general counsel also noted that platforms may feel forced to remove legitimate negative reviews, potentially depriving users of helpful information [15]. Critics fear that because tech companies are held liable for content circulated on their platforms, they will regulate potentially controversial content more strictly, resulting in increased moderation and the removal of opinions across the political spectrum [16]. This removal of opinions, some argue, can infringe on users’ freedom of speech.

An important distinction is that all internet providers are protected under §230, including independent bloggers. Thus, these bloggers and smaller social media platforms could be subject to the same litigation as big tech companies. Should small startups lose their §230 protections, they could face costly lawsuits that drive them offline [17]. 

Another possible implication is that the lack of a federal standard for internet moderation following §230’s repeal could lead to states devising their own laws to moderate social media platforms. This repeal has the potential to place internet moderation power in state hands, which would allow states to either pass laws that limit moderation to preserve free speech or impose stricter moderation standards. This shift in governing power could make it difficult for internet providers to establish consistent moderation policies, a shift that has the potential to hurt smaller start-up social media platforms [18].

The Future of §230

Although a full repeal of §230 seems unlikely, given the failures of past repeal efforts and concerns that a full repeal would lead to the loss of freedom of speech for all users of internet platforms, it is more plausible for policymakers to advocate refinements to § 230's language. One possible refinement is to reform §230 to offer explicit restrictions on generative AI and algorithmic content, since these terms are not clearly defined in §230’s writing. One particular challenge with generative AI is that there is no clear speaker, such as a person or a user on a platform, since technology companies use algorithms to create this content without a person explicitly producing the content [19]. Similarly, TikTok’s For You page and similar features on other platforms are not explicitly controlled by one person [20]. The Center for Democracy and Technology argues that when involved in the creation of new content—even in response to a user prompt—the generative AI system is more likely acting as an information content provider, so future revisions to §230 could potentially identify social media companies as being legally liable for generative AI content [21]. 

One effort proposed in California to impose legal liability on tech companies for deploying algorithms that amplified "willful and reckless content” towards historically vulnerable and marginalized groups is SB 771 [22]. However, Governor Gavin Newsom vetoed the bill due to concerns over the large penalties that tech companies could incur (up to $1 million) and a desire to rely on existing civil rights laws that could meaningfully address violations [23]. Some concern around this bill came from the ambiguous language used in defining “willful and reckless”; refining the language of §230 to explicitly define the extent of content to include explicit threats of violence could help ensure lawful content is not censored. 

What Could an Amendment to §230 Look Like?

For §230 as it currently stands, third-party immunity for tech companies from lawsuits applies in the sense that tech companies are not treated as the publisher or speaker of information provided by users. The algorithmic nature of social media platforms serves as a sort of retweet function for tech companies as they decide which content is most appropriate for a user. Thus, companies could face liability for consistently amplifying content reported as harmful. These efforts could prevent algorithm abuse, such as Facebook’s newsfeed team’s 2018 decision to shift priorities “from societal good to individual value” and take a hands-off approach to regulating hostile content [24].

Additionally, the “good faith” provision of §230 lacks specific language defining what counts as “good faith” behavior, especially given the growing presence of generative AI on social media [25]. The EU AI Act requires “risk assessments, implementing appropriate safeguards, establishing clear oversight mechanisms, and regularly reviewing AI performance to identify and mitigate potential risks” for a tech company to be seen as acting in good faith, a concept that could be applied to refining §230 [26].

However, the complete repeal of §230 has the potential to increase the volume and cost of lawsuits, which could lead to the destruction of small social media platforms and startups. Increased regulatory and financial burden could deter new startups from entering the social media market. Entrepreneurs might be less likely to launch new platforms or services if the risk of litigation looms over operations, leading to greater monopolistic competition [27]. Thus, amending §230 rather than completely repealing it can ensure stronger protections against obscene content while continuing to protect small providers. 

Given the recent Los Angeles Superior Court ruling and past revisions to §230—such as the 2018 FOSTA-SESTA Act, which stripped immunity for platforms that knowingly assist in, facilitate, or support sex trafficking—the door is open to future amendments to §230 that offer stronger protections against violent or inflammatory algorithms and generative AI content [28]. Ultimately, given the recent changes in algorithmic and AI content on social media, it is imperative that lawmakers seriously consider revisions to §230 to better reflect social media’s present state. Although §230 was enacted in the 1990s to protect small internet providers from lawsuits that could stifle internet growth, it is time to revise §230 to better serve internet consumers’ needs in the 2020s and beyond.