es 
Venezuela may be one step away from becoming one of the first countries outside the People’s Republic of China to adopt the new generation of Chinese artificial intelligence systems for mass surveillance and political control, according to a report from the Australian Strategic Policy Institute (ASPI).
The report reveals that in July 2025, the Venezuelan government signed a memorandum of understanding with iFlytek, a Chinese company that has been on the U.S. trade sanctions list since 2019 for its role in the repression of Muslim minorities in Xinjiang. The agreement was negotiated by Delcy Rodríguez, then vice president and now Venezuela’s interim president.
An Infrastructure that was Already in Place
The report, titled “Warning Signals: Venezuela and the Risk of Chinese AI-Enabled Digital Authoritarianism”, documents that Venezuela doesn’t need to build a surveillance system from scratch — it already has one, built largely on Chinese technology.
Researchers identified six interconnected technological layers: telecommunications, digital identity, central databases, video surveillance, internet control, and now, analytics and artificial intelligence. All of them were built or enhanced with support from companies such as Huawei, ZTE, CEIEC, Hikvision, and Dahua.
At the core of this system’s identity layer is the Carnet de la Patria (Homeland Card), a digital ID linked to social benefits. The report states that it had 21.8 million registered users as of January 2023 — roughly three-quarters of Venezuela’s population — and served as many as 14.2 million people each month. Add to this the video surveillance networks from the “safe city” projects and the internet control mechanisms operated through the state telecom company Cantv.
That layer of identity is not the only one. The report also describes a layer of central databases that cross-reference identity information, biometric data, and social welfare data; a network of video surveillance and command centers (such as the VEN911 and SIMA systems) capable of monitoring public spaces and coordinating responses in real time and a telecommunications infrastructure that, by channeling a large portion of the country’s data traffic, functions as an additional checkpoint.
None of these systems need to operate as a single, perfectly integrated system; even when only partially connected, they already enable government agencies to identify, monitor, and take action regarding citizens’ behavior.
What Changes With the Integration of AI?
According to the report, the risk isn’t that Venezuela will replicate China’s comprehensive surveillance system, but rather that artificial intelligence will make the infrastructure that already exists faster, more scalable, and harder to evade.
Facial recognition, voice identification, and automated database cross-referencing tools could be used to identify, locate, and profile individuals without any visible human action, such as attending a protest.
iFlytek, the Chinese company with which Venezuela signed the agreement, is one of the world’s leading providers of voice recognition and content moderation technology. The same report explains that the company promotes a product capable of detecting “more than 100 types” of politically sensitive content in voice messages, and that its technology is integrated into China’s judicial and censorship system.
iFlytek itself already sells a voiceprint identification service that compares recordings against a database to identify a person among thousands; this is precisely the same technology the company has provided to public security authorities in Xinjiang.
If that capability were combined with the network of cameras and the identity records in the Carnet de la Patria, the risk is that the government would no longer need to rely on someone’s visible participation in a protest to identify them, and could instead preemptively flag individuals based on patterns of association, movement, or activity on social media even before any specific incident occurs.
Added to all this is a pattern of opacity. As has been the case for years with technology agreements between Venezuela and the People’s Republic of China, the details of the memorandum with iFlytek have not been made public in their entirety; this makes it difficult to ascertain the actual scope of the agreement, which is why the report recommends demanding the full publication of these agreements.
Despite this lack of transparency, there is at least one precedent that the government itself has publicly acknowledged. In January 2025, the National Chamber of Popular Self-Government was launched—a command center that centralizes information from thousands of community councils and districts across the country—which Maduro described as operating “with Chinese technology and artificial intelligence.”
The Political Context and the Hope of Closing the Window
The agreement with iFlytek was signed months before Maduro’s capture by U.S. forces in January 2026. Delcy Rodríguez, who negotiated it directly at the time, now serves as the U.S.-backed interim president. However, the report notes that during her tenure, the Venezuelan surveillance apparatus has not been dismantled. This is the same conclusion reached by the UN International Fact-Finding Mission, which documented that the surveillance infrastructure remained in place following the change in government.
The authors argue that, despite this, there is a window of opportunity. The interim government’s current dependence on the U.S. could be used to demand the cancellation of the memorandum with iFlytek before this technological escalation is implemented.
What Can Be Done?
Given the risks that China’s technological cooperation poses to human rights in Venezuela, the report offers seven recommendations:
- Support Venezuelan civil society by increasing funding, protection, and training in digital security for organizations and journalists under threat, and by expediting humanitarian visas for those who are being persecuted.
- Provide free access to AI tools so that civil society can circumvent censorship and protect itself from government surveillance.
- The U.S. should demand that Rodríguez immediately cancel the memorandum with iFlytek, before it is fully integrated into the surveillance systems.
- Have the Secretary of State press Rodríguez to dismantle the current surveillance system, including the “Carnet de la Patria”.
- Plan for liberal reform by drafting privacy laws and oversight mechanisms now, in preparation for when political openness emerges.
- Impose sanctions for abuse of surveillance by expanding penalties to include officials and companies that facilitate digital repression.
- Establish security standards for foreign AI used by governments, including independent evaluation and transparency about the origin of the models.
About the Report
“Warning Signals: Venezuela and the Risk of Chinese AI-Enabled Digital Authoritarianism” was written by Fergus Ryan, Bethany Allen, and Nathan Attrill of ASPI’s Cybersecurity, Technology, and Security Program, along with external contributors Mariví Marín (executive director of ProBox), Shannon van Sant, and analyst Alex Colville. The report was published in September 2026 by the Australian Strategic Policy Institute, with funding from the Human Rights Foundation.
*** This article summarizes and contextualizes the findings of ASPI’s original report. For the full document, visit aspi.org.au.
