Effros: Exposing Section 230
In a column last month, I tried explaining why the Cox copyright case now before the Supreme Court was so important regarding the future regulation and business model for the internet. I said we were potentially looking into the abyss depending on how careful the Supremes were in parsing their upcoming decision of that case.
Well, now we have a budding controversy that will also raise similar legal and legislative questions about the future of the current Section 230 indemnification of internet service and edge platforms, especially search and AI providers. Put very simply, the underlying question is why those services should not be treated as publishers, instead of having no responsibility associated with what is put on them, or whether they can be considered at least contributorily liable for what is created and distributed.
The most recent trigger for all this is the controversy, worldwide, over Elon Musk’s Grok, the AI chatbot that allowed users to undress pictures of non-consenting folks on X. Pictures flooded the web, with the predominant victims exposed to this harassment being women and children.
Understandably, a hue and cry ensued. Several nations banned the Musk platforms and others are investigating what “guardrails” have either been crashed or should be created. xAI has already backed down and said it will no longer provide the offending capabilities, at least to those using them for free. But the debate has now started in earnest: are these companies which uniquely empower users to create objectionable content in any way responsible for the results?
Put another way, are ISPs, and especially social media edge providers, always innocent bystanders? Is that so, even though without the creation or delivery they provide, those offensive (or copyright infringing) actions could not have happened? If the ultimate answer to that is no, and thus they can be held responsible, at least in part, then we are certainly going to be “over the edge,” sliding into the abyss I was talking about. And please note, I don’t necessarily think that’s bad! But it will, in no uncertain terms, change the entire business structure, particularly of social media and the rapidly developing AI business models.
It’s impossible to explore all the implications of the sudden surge of discussion about either putting guardrails around such things as the new AI chat capability of sexualizing someone’s picture without their consent, or drafting objectionable (libelous?) material based on instructions followed out by AI. Do the results constitute content that makes the edge provider a “publisher,” and therefore directly responsible or at least secondarily responsible for what is then distributed? If so, then the underlying protection of Section 230 has been (pardon me) stripped away.
That section of the Communications Decency Act served a very valid purpose when it was written: to help accelerate the development of the internet by protecting platform creators. It has clearly worked. But given what we now know are the unintended consequences of that development, do we still need it?
Is it time to tame the internet? Not by the governments, which are inherently burdened with political bias one way or another, trying to write and enforce standards, but rather, allow individuals to bring civil actions against xAI, OpenAI, Google or Meta, etc., for harms caused. There would be a lot of litigation, but it would also force those companies to take responsibility and establish their own guardrails for what they have created. It would be the beginning of a reimagining of how the internet—and particularly social media—should operate. It’s time we did that, while also keeping the government out of it.