China-linked phishing campaigns targeted journalists and activists from sensitive regions like Tibet and Taiwan, using over 100 malicious domains to steal credentials over nine months.
These campaigns, GLITTER CARP and SEQUIN CARP, aimed to support Chinese government interests, suggesting a model of digital transnational repression potentially using independent contractors for plausible deniability.
The broad and relentless nature of GLITTER CARP indicates significant resources and a focus on impact, while SEQUIN CARP, though sophisticated, showed operational weaknesses, potentially leading to future detection improvements.

Atlas AI
Two separate phishing campaigns attributed to China-linked actors targeted journalists and activists over a nine-month period, according to the source material. The operations used more than 100 malicious domains designed to capture login credentials from people connected to Tibet, Taiwan, Hong Kong, and the Uyghur region.
The activity focused on individuals whose work or advocacy intersects with issues viewed as sensitive by the Chinese government. The stated purpose of the credential theft was to enable follow-on operations aligned with Chinese government interests.
Targets included investigative reporters and sensitive beats
Among those targeted were members of an international investigative journalism consortium, along with other reporters covering topics relevant to the Chinese government, the source material said. The campaigns were aimed at gaining access to accounts rather than causing immediate disruption, reflecting a credential-harvesting approach.
The use of a large number of domains indicates an effort to scale outreach and increase the chances of successful compromise across multiple communities and geographies. The source material does not specify how many victims were successfully compromised or what platforms were impersonated.
Two campaigns: GLITTER CARP and SEQUIN CARP
The source describes one campaign, dubbed GLITTER CARP, as broad and relentless. Its pattern suggested substantial resources and an emphasis on impact over concealment, indicating the operators were willing to accept visibility in exchange for reach.
The second campaign, SEQUIN CARP, was described as more focused on journalists and relied on sophisticated personas. Despite that, it showed operational weaknesses, suggesting gaps between the quality of its social engineering and the consistency of its execution.
Contractor-style operations and “plausible deniability”
The activity was presented as consistent with a model in which independent contractors are used for digital transnational repression. In this framing, outsourcing can lower the cost of targeting overseas diaspora communities while creating a layer of plausible deniability for the commissioning state.
The source material links the campaigns’ targeting choices to Chinese government interests, but it does not provide details on contracting arrangements, payment structures, or direct command-and-control relationships. It also does not describe any public response from Chinese authorities.
Operational significance and remaining unknowns
For journalists and civil society groups, credential theft can create downstream risks that extend beyond a single account, including exposure of contacts, research materials, and communications. The source material, however, does not document specific follow-on intrusions tied to these campaigns.
Key uncertainties remain, including the extent of successful credential capture, whether multi-factor authentication limited access, and how long any compromised accounts may have remained under attacker control. The source material also does not indicate whether the malicious domains have been taken down or remain active.


