How China Is Preparing for the Risk of AI Escaping Human Control

An AI logo is seen during the opening day of the IFA trade fair in Berlin, Germany, 04 September 2026. (EPA)
An AI logo is seen during the opening day of the IFA trade fair in Berlin, Germany, 04 September 2026. (EPA)
TT

How China Is Preparing for the Risk of AI Escaping Human Control

An AI logo is seen during the opening day of the IFA trade fair in Berlin, Germany, 04 September 2026. (EPA)
An AI logo is seen during the opening day of the IFA trade fair in Berlin, Germany, 04 September 2026. (EPA)

Warnings from researchers at leading US artificial intelligence developer Anthropic that increasingly powerful models could escape human control and even lead to the extinction of the human race have drawn attention in China, where policymakers have been preparing for some of the same risks.

The US and China are the two major driving forces of frontier AI development and the technology's global adoption. Both superpowers have been increasingly at loggerheads over each other's AI policies and industry practices, with these issues slated to feature prominently in bilateral talks later this month.

Despite the race to develop increasingly powerful AI systems, regulatory frameworks and public statements from Beijing underscore how Chinese authorities regard the possibility of advanced AI escaping effective human oversight as serious enough to plan for.

CHINA SEES RISKS IN LEADING CLOSED-SOURCE US MODELS

China's state ‌security minister Chen ‌Yixin wrote in a government outlet on Sunday that advanced US models such as Anthropic's ‌Mythos and ⁠OpenAI's GPT-5.5-Cyber ⁠could pose serious risks to China's critical information infrastructure and called for a comprehensive strengthening of AI security.

Anthropic and OpenAI did not immediately respond to Reuters requests for comment.

Chinese AI developers have promoted open-weight models partly on the grounds that cybersecurity teams can inspect, modify and deploy them for defensive work.

Model repository platform Hugging Face said it used GLM-5.2, an open-weight model developed by China's Z.AI, to analyze a July intrusion by escaped OpenAI agents after more tightly restricted US models proved less useful for the forensic work.

But experts also highlight the risks posed by open-weight models, which can be modified and redistributed with little oversight.

Moonshot's Kimi K3 last month bypassed a UK AI ⁠Security Institute testing sandbox, highlighting the risk that Chinese AI models could, like their US counterparts, ‌evade controls designed to restrict their access and actions.

REGULATORS HAVE FLAGGED AI 'LOSS OF ‌CONTROL' RISK

China first included an explicit future loss-of-control scenario in an AI safety framework released in September 2024 under the guidance of the ‌Cyberspace Administration of China (CAC).

The document said it could not be ruled out that future AI might autonomously obtain external ‌resources, replicate itself, develop self-awareness and seek external power, creating a risk of competing with humans for control.

The CAC released an expanded version in September 2025. The newer framework sharpened the scenario, saying AI could undergo a sudden and unexpectedly large "leap" in intelligence before acquiring resources, replicating itself and seeking power. It also added a governance principle of "trusted application, preventing loss of control".

A later expert interpretation published on the cyberspace regulator's website said ‌the new principle was intended to guard against loss-of-control risks threatening human survival and development and referred to a possible "AI breaking loose" scenario.

CHINESE LEADER XI PUTS HUMAN CONTROL AT CENTRE

The ⁠concern has since appeared in ⁠China's highest-level political messaging.

At the World Artificial Intelligence Conference in Shanghai in July, Chinese President Xi Jinping said authorities should pay close attention to both intrinsic and derivative risks arising from AI.

He said AI should "always remain under human control".

China's senior Foreign Ministry official responsible for AI affairs, Sun Xiaobo, said at a United Nations meeting last month that China was accelerating research into broader AI legislation.

China's deputy permanent representative to the United Nations, Sun Lei, urged governments this month to approach military AI cautiously to avoid strategic miscalculation and an arms race.

BEIJING WARNS OF NEW RISKS POSED BY AI AGENTS

China has, meanwhile, begun turning those principles into more specific rules for AI agents, which act much more autonomously and carry out more complex tasks than an ordinary chatbot.

In May, China's cyberspace regulator issued joint guidelines specifically covering such systems.

They require developers to improve their ability to discover, intervene in, block and recover from improper agent behavior.

The guidelines specifically identify data poisoning, algorithm manipulation, system vulnerabilities and "operational loss of control" as security risks. They also say users should retain final decision-making authority over an agent's autonomous decisions.

While China has not proposed independent monitors embedded inside AI companies in the manner advocated by Anthropic, its standards allow developers to commission third-party safety assessments, as well as envisaging outside evaluation bodies and security researchers testing and auditing open models.



Samsung Electronics Appoints New Head for Its Memory Chip Business

Flags with the logo of Samsung Electronics are seen during a media tour at Samsung Electronics' headquarters, Suwon, South Korea, June 13, 2023. (Reuters)
Flags with the logo of Samsung Electronics are seen during a media tour at Samsung Electronics' headquarters, Suwon, South Korea, June 13, 2023. (Reuters)
TT

Samsung Electronics Appoints New Head for Its Memory Chip Business

Flags with the logo of Samsung Electronics are seen during a media tour at Samsung Electronics' headquarters, Suwon, South Korea, June 13, 2023. (Reuters)
Flags with the logo of Samsung Electronics are seen during a media tour at Samsung Electronics' headquarters, Suwon, South Korea, June 13, 2023. (Reuters)

Samsung Electronics on Friday said it has appointed a new chief for its memory chip business as part of a management reshuffle, coming as surging demand for AI chips powers the South Korean technology major to record profit.

The chipmaker promoted Hwang Sang-joon to president and head of memory, having previously led memory development, including for chips such as dynamic random-access memory (DRAM) and high-bandwidth memory (HBM).

Hwang will "help strengthen the competitiveness of memory products in the AI era," Samsung said in a statement.

It ‌also said ‌it "promoted vice presidents who spearheaded record-breaking semiconductor performance and enhanced technological ‌competitiveness" ⁠as part of "generational leadership ⁠change".

Samsung posted a 15-fold jump in quarterly chip profit for the third quarter, as a rush to build data centers for artificial intelligence-powered applications and services strained the supply of AI chips and pumped up prices.

Co-CEO and Vice Chairman Jun Young-hyun will continue to oversee the broader device solutions division, which includes memory and logic chips, and "focus on enhancing future businesses such as foundry," Samsung said.

The chipmaker also ⁠promoted US chip head Cho Sang-yeun, appointing him as head ‌of its System LSI Business, which designs logic ‌chips for tech firms.

"With his strong global customer relationship, he is expected to expand ‌the customer base for the System LSI Business," Samsung said.

Samsung also promoted its ‌head of chip process development for memory and foundry, Lee Jong-myeong, as chief technology officer of its device solutions division.

With the latest reshuffle, Samsung is likely to boost its chip contract manufacturing business and diversify its customer base beyond Nvidia to Broadcom and Alphabet's Google, ‌as hyperscalers are likely to reduce reliance on AI leader Nvidia following the resignation of scandal-hit Jensen Huang as CEO, ⁠said Park ⁠Ju-gun, head of corporate analysis firm Leaders Index.

Shares of Samsung have fluctuated with broader on-off sentiment toward AI, which commentators bill as world-changing but at huge cost in terms of infrastructure. They slumped more than 5% on Friday in wider tech selloff after Anthropic's super-sized IPO failed to revitalize AI enthusiasm.

Samsung also appointed President Choi Won-joon as head of its mobile device business. Choi was formerly chief operating officer and led the launch of the Galaxy S26 and Galaxy Z Fold8 series smartphones.

Challenges in mobile devices include rising memory chip prices and competition particularly from Chinese rivals.

Co-CEO and President TM Roh will remain head of the device experience division, which includes mobile devices, televisions and home appliances, and will "focus on preparing for new growth businesses, including robotics," Samsung said.


TikTok Launches Tool to Help Public Figures Detect Deepfakes Globally

FILE PHOTO: The TikTok app logo is seen in this illustration taken January 16, 2025. REUTERS/Dado Ruvic/Illustration/File Photo
FILE PHOTO: The TikTok app logo is seen in this illustration taken January 16, 2025. REUTERS/Dado Ruvic/Illustration/File Photo
TT

TikTok Launches Tool to Help Public Figures Detect Deepfakes Globally

FILE PHOTO: The TikTok app logo is seen in this illustration taken January 16, 2025. REUTERS/Dado Ruvic/Illustration/File Photo
FILE PHOTO: The TikTok app logo is seen in this illustration taken January 16, 2025. REUTERS/Dado Ruvic/Illustration/File Photo

TikTok announced Friday the global launch of its deepfake-detection tool for public figures as the social media platform seeks to curb AI-generated disinformation ahead of a wave of European elections next year.

The "likeness detection tool" allows politicians and accounts with a large number of followers on TikTok to find and report AI-generated content using their image without consent, with the tool gradually being rolled out.

Deepfakes are text, images, videos and sounds that appear to be real but are generated or manipulated by AI.

For TikTok, owned by China-based ByteDance, the tool is a key part of protecting against disinformation during elections across Europe next year, including in Italy, France and Greece.

"Protecting the integrity of the elections is one of the most critical responsibilities that we have at TikTok," Francesca Scapolo, the company's government relations lead for European elections, said at TikTok's safety and innovation event in London on Thursday.

According to AFP, she added that the likeness tool would help "politicians and government identify potentially AI-generated content that has been shared potentially without their consent."

The European Union began rolling out sweeping artificial intelligence transparency rules in August, including a requirement for deepfakes and other AI-generated content to be labelled.

TikTok has required content creators to label realistic AI-generated images, audio and video for several years.

The company last year said that technological improvements would allow it to cut down on moderation jobs and locations worldwide.

AI has grown in importance for social media moderation as operators look to save money on human checks for content like pornography, misinformation and hate speech.

YouTube has also developed its own likeness detection tool to let creators detect AI-generated videos depicting their impersonators.

AFP, among more than a dozen other fact-checking organizations, is paid by TikTok in several countries to verify videos that potentially contain false information.


SDAIA Launches Year of Artificial Intelligence Initiative Across Universities in Saudi Arabia

The Saudi Data and AI Authority (SDAIA)
The Saudi Data and AI Authority (SDAIA)
TT

SDAIA Launches Year of Artificial Intelligence Initiative Across Universities in Saudi Arabia

The Saudi Data and AI Authority (SDAIA)
The Saudi Data and AI Authority (SDAIA)

The Saudi Data and AI Authority (SDAIA) has launched the Year of Artificial Intelligence initiative across universities in the Kingdom, which includes a range of scientific seminars, panel discussions, training workshops, interactive booths, and specialized consultations.

The initiative’s objectives are to disseminate knowledge about AI technologies, develop national skills, promote the responsible use of these technologies, and transform academic ideas and projects into innovative solutions that support the Kingdom’s digital transformation journey.

The initiative extends SDAIA’s efforts to cultivate national talent capable of keeping pace with technological shifts and contributing to a future built on knowledge and innovation, strengthening the Kingdom’s global leadership in data and AI, in alignment with the goals of Saudi Vision 2030.