A fabricated quote becomes a trademark claim
The Atlantic, Forbes and the Guardian are suing Cohere over AI outputs that invent quotes and credit them to real mastheads.
← All tracks
Track
Influence
How information, platforms and persuasion shape what people believe and how they act.
Essays & snapshots
New York makes advertisers confess when an actor isn't real
New York's new law forces ads to disclose AI-generated performers, with fines rising to $5,000 for repeat offenders.
Yandex strikes: did they cut Russian disinformation?
After Ukrainian drones hit three Yandex data centres, officials and commentators said Russian disinformation had fallen in Ukraine, Europe, the US and the UK. The evidence published so far does not support that, and the traffic chart behind the claim measured something else.
Cockroach Janta Party: built online, tracked offline
The Cockroach Janta Party reached 22 million Instagram followers in about two weeks, then saw Delhi Police seal its protest site and detain its founders after tracking them for days, according to police officers. The platforms that made the movement cheap to build also made it visible, and the government can order those platforms to restrict access in ways the movement cannot.
The Pentagon can order an execution. Can it order YouTube to show it?
The Pentagon says it will livestream a military execution on 3 December, but has not said where. The companies that spent a decade removing execution videos from their platforms may refuse to carry it, and the US government may discover how little of the pipeline between camera and audience it actually controls.
The AI election has no control group
American campaigns have adopted generative AI faster than almost any tool in modern politics. Laboratory experiments show it can persuade attentive voters, while decades of field research suggest that campaign contact rarely produces measurable changes in candidate choice. The midterms will be read as a verdict on AI, though they were not designed to deliver one.
Hong Kong's deepfake toolkit is a guide, not a law
Hong Kong's privacy regulator issued a deepfakes toolkit for schools, leaning on existing statutes rather than new rules for the tools themselves.
Meta's own numbers on scam ads say more than its denials
A Reuters investigation found Meta capped anti-fraud enforcement to limit revenue loss, even as it projected scam and banned-goods advertising could generate close to $16bn a year.
Who gets to say what the machines did
Anthropic has published the most detailed public account yet of how artificial intelligence is being used for espionage, propaganda and fraud. It found the activity in its own systems, named the actors, imposed the penalty and released the evidence. The question is what happens when a record like this can only be assembled by the company whose product generated the telemetry.
TikTok's American algorithm is a retrain, not a rewrite
TikTok's US divestiture has closed, but the deal's real test is whether retraining its algorithm on American data changes anything users see.
A marketing technique goes to war
A sanctioned Russian operation optimised its websites for machines rather than people. Five major AI models cited it. The techniques were borrowed wholesale from the marketing industry, and the proposed defence is that democracies borrow them too.
Written for the machines
Generative engine optimisation is a legal, well-capitalised industry that sells influence over what AI systems tell you. The academic evidence says the techniques work. Fraudsters and blue-chip PR firms worked that out before most marketers did.
The FIFA World Cup was a great showcase for Africa
The FIFA World Cup was a great showcase for Africa
The first AI-run cyberattack arrives with an asterisk
Anthropic says a state-linked group used its AI to automate an espionage campaign, but the claim is as much a marketing event as a security one.