At the Gikii conference in 2018, a speaker told us of her disquiet after receiving a warning from Tumblr that she had replied to several messages posted there by a Russian bot. After inspecting the relevant thread, her conclusion was that this bot’s postings were designed to increase the existing divisions within her community. There would, she warned, be a lot more of this.
We’ve seen confirming evidence over the years since. This week provided even more when X turned on location identification for all accounts, whether they wanted it or not. The result has been, as Jason Koebler writes at 404 Media, to expose the true locations of accounts purporting to be American, posting on political matters. A large portion of the accounts behind viral posts designed to exacerbate tensions are being run by people in countries like Bangladesh, Vietnam, India, Cambodia, and Russia, among others, with generative AI acting as an accelerant.
Unlike the speaker we began with, in his analysis, Koebler finds that the intention behind most of this is not to stir up divisions but simply to make money from an automated ecosystem that makes it easy. The US is the main target simply because it’s the most lucrative market. He also points out that while X’s new feature has led people to talk about it, the similar feature that has long existed on Facebook and YouTube has never led to change because, he writes, “social media companies do not give a fuck about this”. Cue the Upton Sinclair quote: “It is difficult to get a man to understand something when his Salary depends upon his not understanding it”
The incident reminded that this type of fraud in general seems to be endemic, especially in the online advertising ecosystem. In March, Portsmouth senior lecturer Karen Middleton submitted evidence (PDF) to a UK Parliamentary Select Committee Inquiry arguing that the advertising ecosystem urgently needs regulatory attention as a threat to information integrity. At the Financial Times, Martin Wolf thinks that users should be able to sue the platforms for reimbursement when they are tricked by fraudulent ads – a model that might work for fraudulent ads that cause quantifiable harm but not for those that cause wider, less tangible, social harm. Wolf cites a Reuters report from Jeff Horwitz, who analyzes internal Facebook documents to find that the company itself expected 10% of its 2024 revenues – $16 billion – to come from ads for scams and banned goods.
Search Engine Land, citing Juniper Research, estimated in 2023 that $84 billion in advertising spend would be lost to ad fraud that year, and predicted a rise to $172 billion by 2028. Spider Labs estimates 2024 losses at over $37.7 billion, based on traffic data it’s analyzed through its fraud prevention tool, and 2025 losses at $41.4 billion. For context, DataReportal puts global online ad revenue at close to $790.3 billion in 2024. Also for comparison, Adblock Tester estimated last week that ad blockers cut publishers’ advertising revenues on average by 25% in 2023, costing them up to $50 billion a year.
If Koebler is correct in his assessment, until or unless advertisers rebel the incentives are misplaced and change will not happen.
Enforcement of the Online Safety Act has continued to develop since it came into force in July. This week, Substack became the latest to announce it would implement age verification for whatever content it deems to be potentially harmful. Paid subscribers are exempt on the basis that they have signed up with credit cards, which are unavailable in the UK to those under 18.
In October, we noted the arrival of a lawsuit against Ofcom brought in US courts by 4Chan and Kiwi Farms. The lawyer’s name, Preston Byrne, sounded familiar; I now remember he talked bitcoin at the 2015 Tomorrow’s Transaction Forum.
James Titcomb writes at the Daily Telegraph that Ofcom’s lawyers have told the US court that it is a public regulatory authority and therefore has “sovereign immunity”. The lawsuit contends that Ofcom is run as a “commercial enterprise” and therefore doesn’t get to claim sovereign immunity. Plus: the First Amendment.
Meanwhile, with age verification spreading to Australia and the EU, on X Byrne is advocating that US states enact foreign censorship shield laws. One state – Wyoming – has already introduced one. The draft GRANITE Act was filed on November 19. Among other provisions, the law would permit US citizens who have been threatened with fines to demand three times the amount in damages – potentially billions for a company like Meta, which can be fined up to 10% of global revenue under various UK and EU laws. That clause would have to pass the US Congress. In the current mood, it might; in July in a report the House of Representatives Judiciary Committee called the EU’s Digital Services Act a foreign censorship threat.
It’s hard to know how – or when – this will end. In 1990s debates, many imagined that the competition to enforce national standards for speech across the world would lead either to unrestricted free speech or to a “least common denominator” regime in which the most restrictive laws applied everywhere. Byrne’s battle isn’t about that; it’s about who gets to decide.
Illustrations: A wild turkey strutting (by Frank Schulenberg at Wikimedia). Happy Thanksgiving!
_strutting.jpg){kind=link}
Also this week:
At Plutopia, we interview Jennifer Granick, surveillance and cybersecurity counsel at ACLU.
Wendy M. Grossman is an award-winning journalist. Her Web site has an extensive archive of her books, articles, and music, and an archive of earlier columns in this series. She is a contributing editor for the Plutopia News Network podcast. Follow on Mastodon or Bluesky.
