cd /news/artificial-intelligence/anthropic-says-it-disrupted-an-ai-po… · home topics artificial-intelligence article
[ARTICLE · art-126834] src=independent.co.uk ↗ pub= topic=artificial-intelligence verified=true sentiment=↓ negative

Anthropic says it disrupted an AI-powered election manipulation platform targeting Malaysia

Anthropic said its Threat Intelligence team identified and disrupted a commercial election-manipulation platform that used its Claude model to target voters in Malaysia, relying on roughly 1,000 fake X accounts, a fabricated news outlet called Malaysia Pulse, and false intelligence dossiers. The operation covered all 222 parliamentary constituencies and used millions of voter records to exploit racial, religious, and monarchical fault lines, and Anthropic said it also observed a request for one million artificial views in support of sitting Malaysian prime minister Anwar Ibrahim, though it found no link between the operation and Ibrahim or his office. Anthropic rated the campaign a Category Two on the Breakout Scale, meaning assets were distributed across multiple platforms without evidence of breakout.

by read3 min views2 publishedSep 11, 2026
Anthropic says it disrupted an AI-powered election manipulation platform targeting Malaysia
Image: Independent (auto-discovered)

Company claims operation used more than 1,000 fake X accounts, a bogus news site, and fabricated intelligence dossiers to target Malaysian voters

  • Bookmark

The campaign, the US artificial intelligence company said in a report, relied on about 1,000 fake social media accounts, a fabricated news outlet, and false intelligence dossiers.

According to the company, “over the past eight months, our Threat Intelligence team identified and disrupted operations in which threat actors tried to use Claude for malicious activity”.

In Malaysia, it said it “identified and removed an account that used Claude to run a commercial election manipulation platform that primarily targeted users in Malaysia”.

It said: “The platform posed as a defensive cyber intelligence and counter-disinformation tooling outlet.”

The operation reportedly covered all 222 parliamentary constituencies and used millions of voter records to identify political and social fault lines, including race, religion and the monarchy, according to one of the key findings of the report.

More than 1,000 fake X accounts were managed through the platform, with mechanisms designed to make them appear genuine and avoid detection by repeatedly changing cookies and IP addresses, Anthropic said.

The company said: “According to the threat actor’s own documentation, the infrastructure was marketed as ‘military-grade, AI-driven, real-time political operations ecosystem’.”

Anthropic’s report comes amid heightened scrutiny surrounding the risks of artificial intelligence, as well as the industry’s attitude to potential threats. Earlier this week, an Anthropic researcher quit the company over fears that the company and its rivals are building AI systems that could wipe out humanity by the end of the decade.

Anthropic said it observed a request “in support of the sitting Malaysian prime minister for one million artificial views on his account”.

However, the company said it had not established who made the request or any link between the operation and prime minister Anwar Ibrahim or his office.

“The operation involved the actors leveraging that platform built through Claude to profile and target voters in Malaysia, constituency by constituency, using census and electoral data that they had ingested.

“The platform managed a network of about a thousand fake accounts that were optimised to inflate engagement metrics and evade social media platforms detection systems. The setup also ran a fake news site called ‘Malaysia Pulse’ by feeding the synthetic news outlet with an AI rewriting pipeline,” it said.

The campaign operated a fake news website called Malaysia Pulse which used AI to rewrite genuine Malaysian news reports and publish them under invented names.

It also reproduced content from Chinese and Russian state-linked media while removing references that could reveal those affiliations. Anthropic said Claude was additionally used to create fabricated intelligence reports containing false claims about an opposition politician and civil society groups.

“These allegations were entirely made up by the actors,” Anthropic said in the report. “We rate this campaign as a Category Two on the Breakout Scale, meaning that the assets were distributed across multiple platforms, but without evidence of breakout into authentic communities,” it said.

The company also said the operator had sought a contract with Malaysia’s national communications regulator, but “we found no evidence that this pursuit succeeded”.

Anthropic said in the report: “Claude refused or partially refused the actor’s requests at several points, including after it had identified a fabricated dossier as material for political defamation and balked at language that explicitly evoked a psychological operation.”

Join our commenting forum #

Join thought-provoking conversations, follow other Independent readers and see their replies

Comments

── more in #artificial-intelligence 4 stories · sorted by recency
── more on @anthropic 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
Live at https://your-agent.zahid.host
Get free account → Pricing
from €0/mo · no card required
LIVE [news/anthropic-says-it-di…] indexed:0 read:3min 2026-09-11 ·