The online safety build up

Here’s what you need to know this week:

• Affairs of state: The Online Safety Bill is reaching final form after five years

State-by-state:

• Apple might be entering the car business, again
• Microsoft is excited about hybrid work
• Amazon’s health business may be a “significant disruptor”
• Google settled a union matter
• Tencent went unnoticed by an MP
• Meta was accused of “extremist activity” by Russia

In one sense, the Online Safety Bill (OSB) is a pioneering piece of legislation: a British attempt to get on the front foot of internet regulation where so many other countries have failed. On the other hand, the bill’s terminology remains ill-defined and unstructured, leading to criticism from both sides of a long-standing debate. Free speech advocates like David Davis MP worry that it represents an unjustified challenge to liberty. Internet-safety advocates worry that it doesn’t go far enough.

Part of the acrimony is down to people not trusting Nadine Dorries, the UK digital secretary with responsibility for the OSB’s passage through parliament. Last month, Dorries arrived at a meeting with Microsoft and asked when they were going to “get rid of algorithms”. 

Essentially, the OSB places a duty of care on internet companies to protect users from dangerous or harmful content. The stringency of the rules depends on two things: the number of users on the platform and whether those users are adults or children. 

Larger “Category 1” services – like YouTube and Facebook – will have to “swiftly take down” not only illegal content, such as terrorist propaganda, but “harmful” content as well. It’s not yet clear how swiftly platforms will have to act and the definition of what makes a “Category 1” platform hasn’t yet been given. Exactly what constitutes “harmful” content is also ambiguous: so far the government has defined it only as material which could cause “physical or psychological harm”. 

To add to the confusion, the latest draft of the Bill – published on 17 March – contains some significant changes to previous iterations: 

• More power to parliament. While previously it would have been up to tech companies to determine what material on their site was “harmful”, now it will be for the government, with the approval of parliament, to determine what content meets this threshold. This change was billed as an attempt to protect freedom of expression and reduce the likelihood that companies over-censor content. But is giving ministers this power any better? David Davis MP argues that ministers will have the ability to introduce “censorship by the backdoor”. The bill gives them the power to define “harmful” content by secondary legislation – allowing ministers to quickly change what “harms” must be removed, with little notice or scrutiny. 

• Combatting internet anonymity. The latest bill contains a new obligation on social media companies to allow users to filter out content from “non-verified” users. Dorries believes a lot of internet harms are caused by anonymous “troll” accounts and this provision gives the government the right to force internet companies to verify users. We don’t know anything yet about how such a verification process could work (the bill says that the “verification process may be of any kind”) – so it could be anything from requiring an image of your passport to entering a simple Captcha – but if it is implemented it could change the internet in a profound way. 

• New offences for communication. The bill adds new offences, including sending “harm-based, false and threatening” messages. And it penalises anyone who causes “intended psychological harm” rather than “requiring proof that harm was caused”. The focus on intent, rather than actual harm, means that “if an individual posted on social media encouraging people to inject antiseptic to cure themselves of coronavirus, a court would have to prove that the individual knew this was not true before posting it” according to the UK government website.

So what does the bill actually stand to do? In general, it’s hard to be clear at this stage how the bill will change social media in practice. This is because the bill leaves a lot up to secondary legislation, and to Ofcom and other bodies to determine how platforms will have to comply. 

Though it’s not yet clear what ministers will define as “harmful” content, there is significant emphasis on removing content which encourages people to self-harm. Other examples of legal and harmful content are likely to include exposure to eating disorders and online bullying. 

What is less clear is the impact of the OSB on the tech states themselves. Ben Greenstone, now of Taso Advisory, was at DCMS when its white paper on online harms was drafted. He told us that big platforms, “like the Metas of the world, will have their businesses entrenched, not harmed, by this legislation because duties [set out by the bill] will be expensive, and the world’s largest tech businesses will be more easily able to comply”. Although Category 1 companies will supposedly be subject to the strictest and most onerous restrictions, all “user-to-user” service providers will be included in the bill. 

“The law is meant to create certainty and predictability, but this bill creates confusion and doubt,” says Robin Wilton, a director at the Internet Society. “The fact that the bill kicks so many issues down the road means that, ultimately, we still can’t tell what its full remit will be. That’s an extraordinarily bad position to be in, given how long the government has been working on the draft.”

Is Apple finally entering the auto sector? The tech state has held discussions with Porsche about “potential joint projects”, according to Porsche boss Oliver Blume. The sports-car manufacturer already puts Apple CarPlay into its vehicles, and Blume said he wanted to “expand on that”. Any deal seems to be in early stages – but it’s consistent with an increasingly close relationship between tech giants and car companies. However, Apple has been exploring how to go beyond CarPlay since 2014, including an abortive plan to design and build its own car. 

Next month Microsoft will reveal an “exciting” (its words) event entitled “Hybrid Work, Powered by Windows”. Of all the tech states, Microsoft is throwing the most weight behind the idea of hybrid work and adapting software for what Julia Hobsbawm, the British writer, calls “the Nowhere Office”. As Satya Nadella, Microsoft’s CEO has said, hybrid work represents “the biggest shift to how we work in our generation”. Microsoft’s hybrid tools, with names like Teams Connect, Neat, and Yealink, sound dull, but they might be as important as Office for many companies in the year ahead. 

Amazon wants to sell you drugs. It also wants to connect you with doctors, deliver diagnostic insights, and provide “on-demand telehealth”. Leaked audio, heard by Business Insider, has revealed that Amazon boss Andy Jassy thinks the tech state’s healthcare business can be a “significant disruptor” of the global healthcare industry. Jassy sees the current issues of long waiting times and inconvenience as a problem with the health sector that Amazon can tackle. It may seem like Amazon Care hasn’t made big inroads since launching in September 2019 (its Halo Band hasn’t exactly been a roaring success); but the company has huge advantages in two areas that might help it dominate health in the future: user data and logistics.

In the latest battle between a tech state and a union, Alphabet, Google’s owner, settled with six employees who alleged the company unfairly stifled worker organising. The workers alleged that Google had fired them and others for protesting the company’s cloud technology relationship with US immigration authorities. If it had gone ahead, the case would likely have revealed “juicy” evidence about Project Vivian, Google’s broader effort to stop unionising, according to Laurie Burgess, the workers’ lawyer. But the claimants had grown emotionally exhausted and took a payout instead. 

Tencent is sponsoring Westminster Games Week next week, when MPs will talk to industry leaders about the economic and cultural contributions of the gaming sector. Christian Wakeford, the MP who defected from the Conservatives to Labour, was supposed to be hosting an event on behalf of Tencent. He cancelled after the Guido Fawkes website questioned whether Tencent was an appropriate partner. What was most surprising was that Wakeford said he’d “never heard of” Tencent even though he’s a vice chair for the all party parliamentary group for video games. Tencent is the largest video game publisher in the world, so Wakeford’s comments were a bit like a member of the parliamentary defence committee asking who Lockheed Martin was.

Facebook and Instagram were already banned in Russia. But this week a Moscow court found the company that owns them guilty of “extremist activity”, underlining just how far and how fast Meta’s relationship with Moscow has deteriorated. Russia had previously designated groups such as the Taliban and Islamic State as “extremist”. But it has more recently applied the same label to Alexei Navalny’s Anti-corruption Foundation. What’s most interesting is that the court specifically said the decision does not apply to WhatsApp, apparently due to its “lack of functionality for the public dissemination of information.” But there are real questions about how WhatsApp will operate given that the court put a stop to Meta’s commercial activities. Read more about the implications of the judgement here.

Thanks for reading,

Alexi Mostrous
@AlexiMostrous

Luke Gbedemah
@LukeGbedemah

Correction: In last week’s Tech States Sensemaker, we stated that Meta’s decision to no longer accept ads from Russian advertisers targeting Russian users would likely cost the company $118 billion in yearly revenue. In fact, Meta’s global revenues totalled $118 billion last year.