Who decides what’s obscene? It’s a question that was once easily answered. For theatre, it was the Lord Chamberlain. For literature, the high court. For cinema, the local council. 

None of these bodies could claim to have a definition of obscenity beyond the standard “I know it when I see it”. In the age of the Internet, however, the censor’s green pen is often passed to the algorithm.

Hang ups around women’s bodies persist even in the Internet age

Instagram, famously, is frightened of women’s nipples, to the point where the advisory board of its parent company, Meta, suggested that the algorithm be rewritten in favour of images of breastfeeding.

The Vagina Museum, dedicated to the cultural history of the reproductive organs of more than half of the world’s population, has repeatedly seen innocent social media posts removed by the mathematical formulation. 

“It took weeks to get our TripAdvisor page up, and we kept getting rejected because our name contained the word ‘vagina’,” the curators told the Stylist magazine in 2021. “Recently, Instagram flagged some realistic illustrations of vulvas we’d reposted from a partner under ‘nudity’. It was only when this was picked up in the news that they changed their stance.”

In February this year, an investigation by the Guardian in association with the Pulitzer’s AI Accountability Network showed that algorithms routinely treated images of women’s bodies as sexual in nature, regardless of context. 

The Guardian reported that guides to medical breast examinations were given the highest score for “raciness” by Google’s AI, whilst Microsoft was “82 per cent confident that the image was ‘explicitly sexual in nature’”. Amazon classified an image released by the US National Cancer Institute as representing “explicit nudity”.

The hang ups around women’s bodies and women’s sexuality, which saw Britain ban some of the greatest books of the 20th century, seem to persist even in the Internet age.

One could question whether a male-dominated industry like Big Tech is equipped to think of women’s bodies as anything other than receptacles of desire or disgust — if, as Caroline Criado Perez the author of Invisible Women might ask, they actually think about women, their bodies or their needs at all.

Blackmailers could exploit vulnerabilities built into apps at the behest of the British government

With that said, it’s hard to argue against the rationale of the Online Safety Bill. Few people think that “online harms” should be ignored, that illegal content should remain online, or that Internet scammers should be allowed roam free.

The problem with the Online Safety Bill is that it is riddled with ill-defined clauses. Platforms are incentivised to remove content quickly for fear of significant penalties. This will inevitably result in innocuous content, which could include educational or supportive content for vulnerable groups, being taken down.

At the same time, the government is mandating client-side scanning in the name of security. This would oblige messaging apps to screen your messages to check you’re not sharing something you shouldn’t be. 

Cybersecurity experts point out that once you create mechanisms whereby encrypted messaging services such as WhatsApp and Signal are compromised, it’s almost impossible to stop the baddies exploiting that gap, leaving users vulnerable to blackmailers, extortionists and espionage. 

Like measures against “obscenity”, these security issues can have a particular effect on women. In the US, after the repeal of Roe v Wade, activists raised concerns that data from period tracking apps could be used as evidence against women seeking abortions.

“We’re very concerned in a lot of advocacy spaces about what happens when private corporations or the government can gain access to deeply sensitive data about people’s lives and activities,” Lydia X. Z. Brown, a privacy expert, told NPR. It is not difficult to imagine how blackmailers or abusers could exploit vulnerabilities built into apps at the behest of the British government in order to intimidate or harass individual users.

Compounding this further, the Bill lacks a requirement for companies to safeguard digital evidence that could be used to punish online offences against women. It is the worst of both worlds: women’s expression is disproportionately limited online, as well as a victims’ right to access justice.

Online safety is complicated, and there are no easy answers. Before progressing with the Online Safety Bill, the government needs to reconsider how it can enable everyone to use the Internet, freely as well as safely.