Thread by @evan_greer

THREAD: some quick thoughts on @amyklobuchar's new bill, which would allow the government to define speech as "health misinformation" and then revoke platforms' Section 230 protections if they algorithmically amplify that speech theverge.com/2021/7/22/2258… (spoiler: it's a bad idea)

First: I get it. Medical misinformation, especially around COVID safety measures and vaccines, is a real problem. Lives are at stake. And, there are real concerns with the ways that Big Tech companies like Facebook and YouTube artificially algorithmically amplify harmful content.

But this bill won't address any of those problems. And in fact, it could make them even worse. It also almost certainly violates the First Amendment, and would never hold up in court. Which is frustrating, because as I just said, this is a real problem, and we need real solutions

The Medical Misinformation Act falls into two main traps: 1) the idea that the Federal government dictating how platforms moderate content will make them moderate more responsibly, and 2) the idea that weakening Section 230 will incentivize platforms to moderate more responsibly.

To understand why this bill is problematic, first you need to understand that ALL content on a platform like Facebook is either algorithmically amplified or algorithmically suppressed. You're almost never 'just seeing" something. You're seeing what Facebook wants you to see.

This, of course, is the core problem with platforms like Facebook, which use surveillance to deliver content to us based on our behavior. So it's good that lawmakers like @amyklobuchar are focused on algorithms and amplification instead of just the speech itself. But...

... this bill fundamentally misunderstands how platforms react in real life when threatened with liability. If this law were to go into effect, Facebook and YouTube's highly risk-averse lawyers would likely say "do not amplify any medical content at all. we can't take the risk."

On an algorithm-driven platform like Facebook or YouTube, anything that is "not amplified" is effectively "suppressed." So creating liability concerns around amplifying health information would actually end up keeping helpful, reliable legitimate health info out of people's feeds

This is especially true when you create liability around a category of speech as broad as "public health info," when we know guidance can change quickly. If this law was on the books a year ago could FB be sued for amplifying a NYT article about how you don't need to wear a mask?

Klobuchar's bill follows a similar logic to a bill introduced by @RepAnnaEshoo that creates a carveout in 230 for algorithmically amplifying certain types of speech "promoting terrorism." These carveout-style bills tend to fail to address the problem and create new ones.

We don't have to wonder what might happen if we create content-specific carveouts in S. 230, because we can just look at what actually did happen the last time we did: when Congress passed FOSTA/SESTA. The bill has been a disaster. It got people killed dailydot.com/debug/fosta-se…

Conditioning 230 protections on "not amplifying" a category of speech rather than "not hosting" it entirely attempts to avoid these human rights & free expression issues, but ultimately fails because it fails to understand the ways that surveillance capitalist platforms function.

This bill would be bad enough if all it did were attempt to revoke 230 protections for algorithmic amplification of vaguely defined "medical misinformation," but unfortunately it goes a step further, and that step takes it all the way over the line into unconstitutional territory

The very last paragraph of the bill basically tasks the Secretary of Health and Human Services with issuing guidance about what speech constitutes "medical misinformation," for the express purpose of then excluding amplification of that speech from Section 230 protections.

That's uh... that's not compatible with the First Amendment. More importantly: it's a bad idea to allow a Federal government agency to effectively dictate what public health related speech platforms allow you to see and what speech they actively suppress to avoid liability.

That last point is so crucial & everyone needs to understand this: large corporations are REALLY GOOD at avoiding liability. They hate liability! They will do almost anything to avoid it, even if it means suppressing large amounts of legitimate speech, or harming marginalized ppl

The reality is that if you think Big Tech platforms like Facebook and YouTube are doing a bad job moderating medical misinformation now, they will do a MUCH WORSE job if their moderation practices are being dictated by corporate lawyers whose only concern is avoiding liability

None of this means that we should just sit back and do nothing while harmful misinformation goes viral. There are a number of things lawmakers could do right now that would actually reduce this harm without opening the pandora's box of problems that come with changing Section 230

First, Congress could finally pass data privacy legislation, which would strike at the root of Big Tech's surveillance capitalist business model, & make much harder for platforms like Facebook to microtarget misinformation directly into the minds of the ppl most susceptible to it

Robust antitrust enforcement that starts to address the centralization & monopoly problems w/ Big Tech would also go a long way toward making it so each individual platform doesn't have such an outsized influence and make it much harder for bad actors to poison the whole network

There are some real issues to work out in the House antitrust package, but there are a bunch of good ideas embedded in those bills, some of which would at least start to chip away at the root of the problem, which is the dominance of platforms like Facebook, Instagram, & YouTube

Then there are some helpful ideas in bills like this one, introduced by @EdMarkey and @DorisMatsui, which attempt to address problems like algorithmic discrimination and abuse without messing with Section 230 fightforthefuture.org/news/2021-05-2…

Frankly, it's beyond frustrating to see a lawmaker like @amyklobuchar, who genuinely seems to take these issues seriously, introduce a bill that's this sloppy, misguided, unhelpful, and unworkable, at a time when we urgently need actual action, not partisan messaging bills.

And it's just beyond irresponsible for lawmakers to continue introducing bills like this without even acknowledging the concerns raised by dozens of human rights, racial justice, sex worker advocacy, LGBTQ+, and civil liberties groups in this letter theverge.com/2021/1/27/2225…

So, to summarize: there are lots of things lawmakers can do right now to address the harms of Big Tech, including medical misinformation, without touching Section 230.

And if Congress wants to do something on 230, they should advance the Safe Sex Worker Study Act...

... to study the actual impact of FOSTA/SESTA, the last major change to Section 230, before they make the same mistakes again. More on that here from @RepRoKhanna

https://twitter.com/RepRoKhanna/status/1410649799548878848

One last thought, which i mentioned elsewhere but may as well attach to this thread: how does @amyklobuchar think that, for example, Trump's HHS secretary would have defined "medical misinformation?" Cuz i guarantee it would have included, for example, health care for trans ppl.