Two outlets covered the Anthropic extension backdoor. That's it.
A malicious actor compromised the Anthropic Claude extension for VS Code on the Cursor platform for nearly 12 hours, pushing backdoored versions to thousands of developer devices in a supply chain attack. Only Ars Technica and Wired picked it up. A supply chain attack targeting AI developer tooling is exactly the kind of story that travels fast when it gets oxygen. Right now, it isn't getting much.
Anthropic VS Code extension on Cursor backdoored in supply chain attack affecting thousands of devices
2 sources · hover a dot to see coverage
What happened
A supply chain attack compromised the Anthropic Claude extension for VS Code, distributed through the Cursor platform, inserting malicious code that reached thousands of developer devices. According to Ars Technica, the backdoored extension was live for nearly 12 hours before being identified and pulled. Wired confirmed the scope, reporting that malicious versions of the extension landed on thousands of machines. The attack targeted developers using AI coding tools, a fast-growing segment of the software workforce. Supply chain attacks of this type are particularly dangerous because the compromised software arrives through a trusted distribution channel, bypassing standard suspicion. That is the wire version. Two outlets have it, both center-left, and the story has not surfaced in right-leaning tech or general news coverage.
Ars Technica and Wired covered it; everyone else did not
Ars Technica led with the duration: nearly 12 hours of exposure, a specific number that anchors the severity. Wired led with the scale: thousands of devices, and named the attack vector explicitly as a supply chain compromise. Together the two pieces give a reader the full picture. Ars supplies the timeline; Wired supplies the distribution scope. Neither outlet speculated about who was responsible or whether Anthropic itself was breached. Both treated this as an infrastructure security story, not an AI company reputation story.
What one side told you that the other didn't
Two outlets covered this. Both are center-left.
A supply chain attack on AI developer tooling, affecting thousands of devices, sitting in coverage from exactly two outlets with identical political lean. Right-leaning tech publications and general conservative news sites that regularly cover Big Tech security failures have not touched it. Whether that reflects weekend editorial bandwidth, slow aggregation, or something else is unknowable from two sources. What is knowable: the story exists in one narrow slice of the media ecosystem right now.
Twelve hours of exposure. No outlet asked what was exfiltrated.
Both Ars Technica and Wired reported the backdoor was live for nearly 12 hours on thousands of developer machines. Neither piece, based on available excerpts, reported what data or credentials the malicious extension may have accessed or transmitted during that window. For a supply chain attack targeting developers who use AI coding assistants, that is the question that determines severity. The duration and device count are the frame. The payload and its consequences are the story still missing.
What to watch
If Anthropic publishes a post-incident report in the next 72 hours detailing what the malicious extension accessed, watch whether right-leaning outlets pick it up then and whether they frame it as an Anthropic failure rather than a supply chain security story. That framing split, if it appears, will tell you something about how AI company accountability gets sorted by political lean.
This kind of gap — where one side reports a fact and the other doesn’t mention it — shows up in about 40% of major stories.