The Notebook: Labour is charting a dangerous course on free speech

Labour’s proposed expansion of the Online Safety Act is a troubling step for free speech, writes Big Brother Watch director Silkie Carlo in today’s notebook

Three policies to watch out for

The new Labour government has started to chart a dangerous course on free speech in Britain, which could roll back free speech rights by decades. As the gears of the new government get into motion, here are three policies in the pipeline to watch out for.

First, senior Labour figures are calling for an expansion of the Online Safety Act, which promised to be a much-needed bill to increase transparency but fast became a censor’s charter, inventing loose concepts like “legal but harmful” speech that must be censored. Ministers recently held discussions with the Centre for Countering Digital Hate on expanding the Act to create alarming new ‘emergency’ online censorship powers.

Second, the new home secretary is planning to expand police recording of so-called ‘non-crime hate incidents’ – a U-turn on the last government which sought to reduce petty non-crime reports. Big Brother Watch’s recent FOI investigation found police are still recording ‘non-crimes’ such as a man putting Israeli flag stickers on lamp posts, and children drawing with chalk on pavements – which can show on individuals’ criminal history checks. The plans are likely to make this pointless recording far worse.

Third, Labour have rebooted the secretive Counter Disinformation Unit (now renamed the ‘National Security Online Information Team’) and tasked it to snoop on online discourse about controversial topics such as Israel and Gaza. Given the Johnson government used the unit to flag his critics, from former Green MP Caroline Lucas to ex-minister Sir David Davis MP, could Starmer’s government task the unit to snoop on opponents of his plans to take winter heating support from thousands of pensioners?

Excessive censorship never benefits the public – it always benefits the powerful.

Why is Labour lagging on AI laws?

This evening, parliament will hear for the first time from victims of AI harms.

Thirty-eight-year-old Londoner Shaun Thompson, who was wrongly stopped and questioned by the Metropolitan Police this year following a misidentification by live facial recognition surveillance cameras, will address parliamentarians in an event about the need for laws to prevent AI harms.

Parliamentarians will also hear from experts who worked on the EU AI Act – which broadly prohibited live facial recognition and other dangerous forms of AI across the continent – who might be wondering why the UK is lagging so far behind. Starmer curiously dropped AI legislation at the last minute from the King’s Speech – some analysts say under pressure from the relentlessly pro-AI, pro-technocracy lobbying outfit, the Tony Blair Institute.

However, with AI already expanding the database state and making mistakes in policing, welfare and marking children’s GCSE results – how many people’s lives have to be impacted before science minister Peter Kyle takes action?

Lord Clement-Jones takes on the mantle

Where the government has failed to control AI, parliamentarians are taking the lead.

Lord Clement-Jones, the Liberal Democrat’s AI spokesman and author of the authoritative guide to AI regulation Living with the Algorithm, put forward a private members’ bill yesterday that would create transparency over AI decisions made by the state.

Transparency over computer decisions affecting people’s lives isn’t much to ask – particularly in the wake of the Horizon scandal. But whether the government backs the bill looks highly uncertain.

Worth listening to

Talking of the erosion of free speech online – to hear first hand from those influencing the Labour government, I recommend Triggernometry’s new interview with Imran Ahmed, a former Labour staffer and now CEO of the Centre for Countering Digital Hate. You can catch the debate on Youtube and Spotify.

Ahmed appears to have such a weak, uninformed framework for thinking about speech policies online, he answers the interviewers’ question about whether content calling for “armed revolution” should be permitted online with a shrug. It’s an extraordinary moment. Why Ahmed struggles with this, yet has produced reports exhibiting Kermit memes as apparent evidence of “dangerous” content online that should be censored, is baffling.