Send to HN

from the getting-it-all-wrong dept

For years now we’ve written about the problems of the UK’s latest (in a long line) of attempts to “Disneyfy” the internet with its Online Safety Bill. While the bill had faced some hurdles along the way, made worse by the ever-rotating Prime Minister position last year, there was talk last week that some more hardline conservatives wanted to jack up the criminal penalties in the bill for social media sites that don’t magically protect the children. And, while new Prime Minister Rishi Sinak had pushed back against this, in the end, he caved in.

Michelle Donelan, the Culture Secretary, has accepted changes to the Online Safety Bill that will make senior managers at tech firms criminally liable for persistent breaches of their duty of care to children.

One of the worst aspects of the bill in earlier forms — possible punishment for legal speech, if deemed harmful — remains out of the bill, but that’s little comfort based on these new criminal additions.

Tech platforms will also have a duty of care to keep children safe online. This will involve preventing children from accessing harmful content and ensuring that age limits on social media platforms – the minimum age is typically 13 – are enforced. Platforms will have to explain in their terms of service how they enforce these age limits and what technology they use to police them.

In relation to both of these duties, tech firms will have to carry out risk assessments detailing the threats their services might pose in terms of illegal content and keeping children safe. They will then have to explain how they will mitigate those threats – for example through human moderators or using artificial intelligence tools – in a process that will be overseen by Ofcom, the communications regulator.

That is, this is more or less California’s terrible law (which we were told was modeled on existing UK law, which was clearly not true if they’re now implementing this new law). Anyhow, the criminal liability part is absolutely ridiculous:

Even before the government conceded to backbench rebels on Monday, tech executives faced the threat of a two-year jail sentence under the legislation, if they hinder an Ofcom investigation or a request for information.

Now, they also face the threat of a two-year jail sentence if they persistently ignore Ofcom enforcement notices telling them they have breached their duty of care to children. In the face of tech company protests about criminal liability, the government is stressing that the new offence will not criminalise executives who have “acted in good faith to comply in a proportionate way” with their duties.

It’s one thing to say we won’t criminalize you for acting in good faith, but it’s another thing to have your freedom on the docket and have to litigate that you acted in good faith. And, these are government officials we’re talking about. They’re not exactly known for acting in “good faith” when demonizing tech companies.

Wikipedia has already expressed reasonably grave concerns about all of this.

Again, there are so many problems with the setup here that it’s difficult to know where to start. First off, as we’ve discussed, the narrative about the internet being harmful to children appears to be massively overstated, and there’s actual evidence that it’s actually helpful to way more children than who find it harmful. That doesn’t mean that we shouldn’t look to reduce the harms that seem to impact some (because of course we should!) but these bills are often written in a way that assumes all harm that comes to children is from social media and that there are no redeeming qualities to social media. And both of those things are false.

Second, if you have to do special protections “for the children,” you’re almost certainly leading to kids being put at even greater risk, because the whole framework forces websites to do age verification, which is a highly intrusive, privacy-diminishing effort that actually is harmful to children in and of itself (i.e., this law almost certainly violates the law, because the authors took no “duty of care” to make sure it protects children).

So, now, all children will be tracked and monitored online, exposing their private information to potential breach and, even worse, teaching them that constant surveillance is the norm.

As for the companies, the risk of not just huge fines, but now criminal liability will mean that all of the incentives are to over-block, and not allow anything even remotely controversial. This is not how you teach children to be good, contributing members of society. It’s how you make it so children are kept in the dark about how the world works, how to make difficult choices, and how to respond when they’re actually put in a dangerous scenario.

It is, again, exactly how the Great Firewall of China initially worked: not by telling service providers what to block, but making it known that any “mistakes” would lead to very strict punishment. The end result is massive over-blocking. And that means all sorts of useful content will get buried. Because if you’re an executive for one of these companies and you’re facing a literal prison sentence if you make the wrong choices, your focus is going to be on being super aggressive in blocking content, even if the actual evidence suggests doing so creates more harm than good.

Just as an example, there are plenty of stories about content being shared about eating disorders. Almost certainly, under the Online Safety Bill, most sites will work to hide all of that content. The problem is that this has been tried… and it backfires. As we covered in a case study, when sites like Instagram did this, kids figured out code language to talk about it all anyway, and (more importantly), it was found that having these groups more open allowed people to better come in and help kids recognize that they had a problem, and to get them help. Simply hiding all of the content doesn’t do that.

Once again, this could mean that kids will be put in greater danger, all because a bunch of prudish, stuffy politicians have no idea how people actually act, or how the internet actually works.

Filed Under: , , , , ,


Read More