
Art direction by Adam Musaev | Art by Google Notebook
AI disclosure: No Artificial Intelligence was used in researching, organizing or writing this article. All multimedia was created with AI. Google Gemini Notebook was used to create the summary, video explainer, and infographic. Google Gemini was used to create all of the political cartoons. Prompts for images are listed at the end.
Summary: AI is rapidly outpacing current legal frameworks, creating urgent challenges regarding internet safety and online speech. The foundational statute governing digital platforms, Section 230, faces severe bipartisan scrutiny and potential dismantling as lawmakers grapple with emerging technologies. Generative tools complicate traditional legal definitions by producing complex synthetic outputs, deepfakes, and controversial media that test decades-old First Amendment protections.
Introduction
Artificial Intelligence (AI) is a contentious social, political, environmental, and technological issue. AI is starting to become integrated into everyday life, and many people use AI daily. More importantly, AI is moving faster than the laws meant to govern it.
Technology CEOs from major companies like X, Nvidia, META, OpenAI, Anthropic and Google are largely in favor of rapid, unrestrained development of this technology, despite recent calls from a Sept. 12 NPR report calling for a slowdown.
Many CEOs from the aforementioned companies feel as though the path and any harm potentially done is inevitable as far as AI integration into everyday life. Sam Altman, the CEO of OpenAI recently said in an interview with Politico that “we believe that the world should accept some bad things happening for the benefits of this technology and people having the agency.”.
According to a 2026 Pew Research Center study, more than half of Americans say they feel more concerned about AI than they are excited. Society is still working through how to measure risk, what counts as speech, how AI will replace jobs, and how to verify what is real. The single federal law that might govern AI speech is Section 230 of the Communications Decency Act (CDA).

Art direction by Adam Musaev | Art by Google Gemini
Defining Section 230
“No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider.” (47 U.S.C. § 230(c)(1))
Section 230 of the Communications Decency Act of 1996 is the main federal law governing internet speech. Section 230, now 30 years old, serves as the 26-word foundation by which the idea that users, not companies, are accountable for speech on platforms such as Twitter, Facebook, or Bluesky. The law also lets companies remove objectionable or illegal content in good faith without punishment.
With AI, many, including Mason Owens, who is with a Trust and Safety Team contracted to a MAANG company, warned in an October 2026 interview that AI outputs “are more of a Frankenstein’s monster” that “just feels like,” speech rather than a model having “independent thoughts or feelings of its own.” The uncertainty of where this can go moving forward makes many nervous.
Pundits and policymakers approach Section 230 from a range of politicized viewpoints. In various articles, liberal commentators and entities, such as the American Civil Liberties Union (ACLU) and the Electronic Frontier Foundation (EFF) defend Section 230 as fundamental to preserving internet freedom and free speech.
The Bipartisan Policy Center’s analysis supports and emphasizes a “polarized framing of pro-content moderation versus anti-censorship”. This division on Section 230 reflects deeper ideological divides in governmental and media culture. These divides also will undoubtedly also have an effect on the AI debate. Politicians in Washington, meanwhile, have much different views on Section 230.
What Washington D.C. Says About 230
Bipartisan criticism exists in both chambers of Congress. U.S. Rep. Frank Pallone (D-New Jersey) in a press release has advocated for sunsetting Section 230 to hold big tech accountable. Former House Speaker Nancy Pelosi (D-California) in a podcast interview later reported by The Hill has referred to Section 230 as a “gift” to tech companies, implying that if platforms didn’t do more to address harms lawmakers could take that gift away.
In an interview with the New York Times Editorial Board, Former President Joe Biden in 2021 suggested that Section 230 should be “revoked” because social media platforms knowingly allow falsehoods to spread. In 2025, Dick Durbin (D-Illinois) argued on a Bluesky post that platforms must stop “addicting kids and enabling drug trafficking on their platforms.”.
On the other side of the aisle, both the first and second Trump administrations have taken aim at Section 230 as well. In 2020, reporting by NBC News said that Trump attempted to repeal Section 230 by threatening to veto the annual defense funding bill.

Art direction by Adam Musaev | Art by Google Gemini
According to NPR, Trump posted to Twitter arguing that Section 230 was a “serious threat to our National Security & Election Integrity,” claiming that America could never be safe if the law were to stand.
Despite President Donald Trump’s claims, more substantive conservative criticism of Section 230 reveals the right’s true framing on the issue. Trump, during a Social Media Summit reported by VOA, argued that tech giants were censoring conservative voices.
The Department of Justice under Bill Barr conducted a review of Section 230, the Federal Communications Commission (FCC) and other governmental bodies have all either called for clarifications or limitations in similar reports.
Project 2025 by the Heritage Foundation is the controversial and comprehensive policy framework for the second Trump administration. Project 2025 features language recommending a narrowing of Section 230 protections. Federal Communications Commission (FCC) chair and co-author of Project 2025, Brendan Carr, has signaled possible reinterpretations of Section 230. The primary goals stated in Project 2025 are:
- Eliminate immunities that courts added to Section 230.
- Impose transparency rules on Big Tech.
- Support legislation that scraps Section 230’s current approach.
- Support efforts to empower consumers.
- Require that Big Tech begin to contribute a fair share.
- Protect America’s National Security.
As of 2026, politicians appear to have gotten their wish for change. A bipartisan effort as described in this press release from Durbin and the late Lindsey Graham (R-South Carolina) to sunset Section 230, is currently sitting in the Senate Committee on Commerce, Science, and Transportation. If this were to be passed into law, advocacy groups, including the EFF, warn of the unintended consequences if the current law is dismantled without an adequate or flushed out replacement.
Section 230 is the central law in ongoing debates about internet responsibility, innovation, and public safety. Any move to repeal a piece of legislation like Section 230 warrants caution. The unintended consequences of a world without Section 230 could mean that new liability rules could undermine free expression and possibly foster the conditions in which state-owned/influenced media drowns out or dominates diverse viewpoints.
Politics aside, what does this mean for the rise of AI?
How AI Fits
Mason Owens works to enforce guidelines to ensure alignment with responsible AI ethics as a contractor with a MAANG company. In an interview on AI and Section 230, Owens was asked about how the Section 230 debate could map out onto future debates regarding generative outputs and AI.
“Roughly speaking, an AI output is an amalgamation of all the training data that it has access to that it thinks would be relevant to an ideal response to the prompt,” Owens said. “It is never creating something from nothing. It is more like it is stealing bits and pieces from so many places that it produces something that feels like something new. It also never creates anything without the input of a user, which further supports the idea that it still remains a platform.”
The courts thus far have ruled erratically when it comes to free speech and AI. There does not seem to be a general consensus either in the courts, the law, or society. The reaction to the Seventh’s Circuit Court decision in United States v. Anderegg where AI child pornography was deemed protected under the First Amendment, reflects the still evolving nature of understanding of what AI-generated content is and is not.
Seventh Circuit Decision
In September 2026, Federal prosecutors in Wisconsin charged a man with “producing, distributing, possessing, and transferring sexually explicit AI-generated depictions of children.”. Anderegg used a generative AI model to make more than 13,000 images of AI-Child Sexual Abuse Material (CSAM). In the ruling, Judge John Lee explained that “under Stanley v. Georgia, individuals have a First Amendment right to possess obscenity in the privacy of their home. 394 U.S. 557 (1969).”

Art direction by Adam Musaev | Art by Google Gemini
Lee ultimately felt bound by Supreme Court precedent protecting possession of pornography, and a later ruling criminalizing CSAM. “Given the relentless advancement in artificial intelligence models, we have some concerns about the lines these cases draw, but we are not free to redraw them ourselves.” This ruling from the 7th Circuit has prompted Judge Lee to encourage the Supreme Court to have “the prerogative of overruling its own decisions.”.
Jessica Schidlow, a Senior Policy Analyst & Legislative Counsel working with Enough Abuse, explained in a blog that “The Supreme Court precedents shaping Anderegg date from 1969, 1990, and 2002. Generative AI has dramatically altered the factual assumptions underlying those decisions.” wrote Schidlow. Given that AI can now produce nearly photorealistic images and videos of people both real and fake, it is easy to understand how the Seventh Circuit Court had such trouble determining legality, given the Court’s framework is functionally antiquated.
Luckily, if this case somehow doesn’t reach the Supreme Court, New York State Senator Patricia Fahy has already taken steps to “Criminalize the Possession of AI-Generated CSAM”. This could lay a framework for other states to follow suit.
In a press release, Fahy’s office said that “Protecting our children crosses party lines and this legislation sends a strong message to those who seek to do them harm or use artificial intelligence to generate obscene material: New York State will hold you accountable under our laws.”

Art direction by Adam Musaev | Art by Google Gemini
AI Actors?
An AI product, named Tilly Norwood, has sparked controversy pertaining to whether AI has speech, personhood, or thoughts of its own. Particle 6 Productions, a UK based “AI-generated studio,” are the creators of AI actress “Tilly Norwood.” Norwood acted in a movie and went on a press tour to promote this product. The question is if this AI generated character has a right to free speech, or thoughts of its own.
Real actors have put their two cents in. Joseph Gordon-Levitt, known for roles in “Inception”, “Snowden” and “500 Days of Summer” weighed in at Internet Governance Forum, where he said:
“When you think of a character like Donald Duck, for example, you don’t say that Donald Duck is an actor. You don’t say that Donald Duck gave a performance. You say that Donald Duck is an animated character that was created by a number of different artists, writers, animators, voice actors, etc., and all of those artists are given credit for creating Donald Duck. The product known as Tilly Norwood is no different.”

Art direction by Adam Musaev | Art by Google Gemini
Owens thinks products like Norwood and the company behind it, can contribute to feeling like “I’m not making a difference. Even if one company invests heavily in safety, who is going to stop another one from renting out an AI-generated mask like ‘Tilly Norwood’ that people can use to mislead others?” said Owens.
This misleading has affected all generations, young and old, and no more has this had a clear effect in recent reporting from NPR where political campaigns have now been flooded with AI-advertisements in the 2026 Midterm Elections.
AI used in this way could have real consequences. Some candidates however, are less worried about AI-generated political ads. Sen. Mike Lee, R-Utah said voters “are smart enough to spot comedy and obvious parody when they see it,” in a Sept. 16 interview with NPR.

Art direction by Adam Musaev | Art by Google Gemini
Moving Forward
With Section 230 in potential jeopardy, and AI innovation accelerating at light speed, Congress and the White House will soon potentially have the opportunity to redefine internet speech. Beyond internet speech, looking at speech through the lens of AI could and should be on the table in proposed reforms.
Where does this leave society, and more importantly American law going forward? Even the “Godfather of AI”, Geoffrey Hinton warned in an interview on Capitol Hill that society may only have “maybe a year, but not much more than a year,” to regulate AI systems.
This strict framing of speech, authorship, and prompting should ultimately reflect the framework seen in Section 230, where prompting AI models is no different than a user using a platform to make speech. The liability ultimately is with the author, but companies should have a legal standard by which to moderate.
Owens concurs: “There needs to be a universal set of safety standards or else there will always be a new model popping up that lacks the restrictions that the others have. This needs to be a conversation that takes into account all the harm AI has caused and all the potential harm that could still result from its unrestricted growth.”

Art direction by Adam Musaev | Art by Google Gemini
Legislators like Durbin gave the Trump administration a golden ticket to reform internet speech. Until the midterms, AI companies according to recent reporting from an event at the White House can functionally do whatever they like when it comes to AI safety. This is on top of major reversals of Biden-era executive orders, according to Associated Press reports.
Courts and lawmakers have not said where free speech protection for AI starts or ends. Many legal experts argue that most AI output counts as protected speech according to Foundation for Individual Rights in Education’s (FIRE) analysis, but no clear line exists for deepfakes, political ads, or chatbot responses.
Trump has said in a Truth Social post that “The only control or ‘guardrails’ that AI needs is a STRONG AND SMART (High IQ!) PRESIDENT, and the U.S.A. has that, in spades!”. That push and insistence on no guardrails comes as the administration has lost 75 of 93 First Amendment cases, according to a Truthout report.

Art direction by Adam Musaev | Art by Google Gemini
Conclusion
The rules and laws written in the next few years could shape both internet speech and AI-generated speech at the same time. Content moderation, the issue Section 230 was built to handle, and AI-generated speech are now becoming the same legal problem.
Congress has not yet decided who is responsible when an AI system produces harmful speech, whether that be the user, the company or no one.
Until lawmakers or courts answer that question, the companies building AI will largely set the rules themselves. Adam Smith warned about this risk in “The Wealth of Nations”, writing that people (merchants) in the same trade rarely meet “but the conversation ends in a conspiracy against the public, or in some contrivance to raise prices.”. Trusting billionaires, who stand to profit, to self-regulate guardrails that serve those ends can be extremely dangerous for the industry and society.
Video direction by Adam Musaev | Video by Google Gemini Notebook

























