Why are Americans so divided? There is a villain here that never gets discussed. It’s called Section 230 of the Communications Decency Act. This was the provision that shielded internet providers, and later social media networks, from liability for the content of users.
Why is this so important? It means people can say whatever crazy, inflammatory, hateful, and factually incorrect things they want and social media platforms don’t have any incentive to stop them. In fact, they can even encourage them!
The reason extremist partisans can so easily find an audience is Section 230. Not only are their views not censored in any way but the algorithm helps you find them!
If TikTok was liable for content that led directly to violence, suicide, self harm, etc. the algorithm would behave a lot differently, wouldn’t it?
This is a lot different from TV, newspapers, etc. If a newspaper ran a story that was slanderous or indecent or in some other way violated the law, it can – and will – be sued. Thus, news organizations have fact checkers, editors, etc. to police their content to avoid potential liability.
If this standard were applied to social media, we wouldn’t have social media. Or at least that’s what Facebook and Twitter want you to believe.
Alternate Reality
But is it true? I suspect, if social media had liability for its content, it would look a lot more like LinkedIn where people have an incentive to police themselves, so as not to antagonize current or future employers.
It wouldn’t be all warm and fuzzy (LinkedIn still can get ugly at times), but it would be something you’d be much more comfortable letting your kids participate in.
What else might be different if social media was liable for content on its platforms?
I suspect there wouldn’t be aliases. People would have to post under their real name and be verified to reduce liability for the platform.
People behave a lot better when they talk in person than online. That’s because people care about how they are viewed by others. That disappears behind phony user names.
There would be a lot, lot, lot less disinformation. And not just about politics. Fewer crazy get rich quick schemes or harmful diet fads or romance scammers. Fake news might still exist, but it wouldn’t be a daily topic of conversation.
One thing that wouldn’t change is cat videos. Algorithms would still push cat videos! But wouldn’t be better off if there were just as many cat videos but far fewer disinformation videos?
Section 230 History
How did we get here? Well, we really weren’t supposed to. The original intent of Section 230 was NOT to give broad immunity to internet providers.
It was supposed to provide them immunity from being sued for removing content! That’s right. The origin story was internet companies were being sued for their decisions to remove content for safety (e.g. blocking porn from minors).
The language in the bill gave protection to internet providers to allow them to remove more content and not get sued.
Unfortunately, later on, a judge opened Pandora’s Box and decided they were also immune if they did not remove offensive content and that began the race to the bottom that we all experience today.
So as much as I’d like to blame Al Gore for the decline of modern society, the truth is it was really some US district court judges who changed what Congress wrote and bear no accountability for their actions.
But What About AI?
Glad you asked. You might think this will only get worse under AI, what with its hallucination and continuous egging on of would be psychopaths.
Well, maybe not.
Remember, Section 230 protects social media companies from liability for content created by individuals. But who creates the content produced by AI?
The AI does!!!
That means AI has no protection for harmful content it produces. And guess what? It produces a lot of harmful content.
When an AI chat leads to someone committing suicide or murdering someone (yes, these things have already happened), then AI is liable.
These cases have obviously not gone through the courts yet, but they will.
You want a bear case on AI? It’s going to be sued out of existence. Good luck, Sam Altman!
The Insurance Opportunity
AI is a huge opportunity for insurers…to lose boatloads of money!!!
I have no idea how big a tower OpenAI or Perplexity buys but Meta and Google are developing their own AIs and they have huge liability towers. Mark Zuckerberg has made very clear he wants his AI to have no guardrails.
If you are an insurer who views these big tech companies as a prestige client, I wish you good luck! You’re going to need it.
If you’re writing the excess at the top of these towers for three pennies per $1M, you might as well just tell S&P to downgrade you now.
If you’re a reinsurer of the types of big primaries who likely lead these big liability towers, yeah I’m sure you have no idea you’re even exposed to this.
And remember, this is just the liability on the AI piece. Imagine if a future Appeals Court decides to restore Section 230 back to its original language and takes away the liability protection on prior years?
Call Your Representative
If you haven’t fainted from the section above, I’m going to return to the main topic which is the best thing Congress can do to heal our country is amend Section 230 to remove immunity from liability for social media platforms.
It would put most of the extremist agitators into early retirement. And if these “content creators” had to find normal jobs, no crazy people would try to shoot them (or worse, innocent children to spread their manifesto of ideas they learned online).
We can return to arguing over less harmful things like sports and reality shows and celebrity gossip…and which cat video is the most entertaining.
We’d have a lot fewer shootings, a lot more of boring politicians who actually accomplish things, and we’d finally be able to talk to our family and neighbors without worrying the conversation would turn to politics.
If you’d like to see this version of America, call your Congressperson and demand they change Section 230. It would be terrible for insurers, but great for America.
In the last section, first lined, bolded, I think you mean “immunity” rather than “liability”.
Thank you, it’s now corrected.
I agree harmful speech online is a real problem, but scrapping Section 230 is a blunt move with big costs. Section 230 lets new platforms moderate content without being treated like publishers. Remove it and all will likely over censor, small services will shut down because they can’t afford the lawsuits, and power will end up with a few big companies or the courts, which will not necessarily make things safer.
A better approach is targeted fixes like requiring algorithmic transparency and audits for recommendation systems, which will better hold platforms accountable when they design or amplify harmful content, label AI generated material, etc.. These steps reduce harm while protecting free speech and innovation.
I do not support exposing people’s real identities or “doxxing” as you will as a way to hold people accountable. Forcing people to use real names harms free speech and puts vulnerable people, reporters, and critics at risk.
The LinkedIn example in the article is frail. Just because someone uses LinkedIn for work does not mean it is fair or safe to treat all their online activity as public. LinkedIn profiles for example can be fake and many people keep separate professional and personal accounts for legitimate reasons.
If accountability is the goal, focus on better evidence standards like platform algorithmic recommendation transparency.
Publicly exposing people is a blunt, harmful tool that will cause collateral damage without fixing the real problems.
“Forcing people to use real names harms free speech”…lol, of course you don’t support this because you are too cowardly to put your name behind your opinions (even the rare reasonable ones like this one) and would rather harass others without accountability.
Protections for anonymous speech are meant to protect, as you say, those who are vulnerable to adverse consequences for their speech. None of that applies to keyboard warriors like you whose only risk is embarrassment once people realize how poorly you act online.
PS: I didn’t delete this one for hiding behind anonymity again since your content was respectful. If you can continue that, you are free to rebut, but if you go off on another of your unhinged rants, it’s getting removed.
“Harass” and “unhinged.” Interesting word choice and a dishonest framing of each, since none of my posts harassed anyone nor were they unhinged in the true sense of either word. Rather, they disagreed with your positions in a very “direct” manner, which you’re clearly sensitive to.
I didn’t write what I did about anonymity by accident since it was done to prove your desire to know my real identity so that you can be “retaliatory”, or as you call it, enforce “accountability” lol.
No “embarrassment” here. That’s just projection due to your sensitivity to my posts and further unmasks why you would like censoring “private” speech. There’s nothing objective in your response to my post, but of course this will likely also be deemed “harassment” and “unhinged” and removed by you as well.
Obsessively responding over and over and over again cause you’re mad at the spam filter is pretty unhinged, especially doing it from multiple accounts and whatever other shenanigans you tried. That is not something a normal person does.
Btw, I don’t care who you are. The issue is you hide behind anonymity to avoid accountability. It is a cowardly way to behave. If you believe something, you should stand behind it, which is what I do with everything I write. Even if I knew who you were, I could no more retaliate against you than you have “retaliated” against me with your online stalking, so you seem to also suffer from paranoia if you think I could harm you somehow by knowing who you are.
The broader point is if you feel so strongly about what I write (for whatever crazy reason you have), then respond to me directly. You can DM me on LinkedIn and I will debate you with far more rope than I will give you here, but this blog isn’t about you so I will most likely delete any further comments unless they are clearly on topic and with an appropriate handle.
Does “corporations are people” approach change the stance that “AI has no protection for harmful content it produces.”? If AI is effectively a voice of a corporation, it is protected. No?
(“Section 230 protects social media companies from liability for content created by individuals.”)