分类： technology

In a high-profile move that highlights growing regulatory scrutiny of cross-border artificial intelligence deals, China’s top economic planning body has formally blocked Meta Platforms’ planned purchase of Manus, a Singapore-headquartered AI startup with founding origins in China. The announcement, made public on Monday, marks one of the most prominent restrictions on a major U.S. technology company’s acquisition of an AI-linked firm with Chinese connections in recent years.

The official order came from the Office of the Working Mechanism for Security Review of Foreign Investment under the National Development and Reform Commission (NDRC), China’s leading planning agency. In its brief public statement, the NDRC confirmed it was prohibiting the foreign takeover of Manus and mandating all parties involved unwind the transaction completely. Notably, the agency did not explicitly name Meta, the American parent company of major social platforms Facebook and Instagram, in its public notification.

This formal ban follows a preliminary investigation launched by Chinese regulators earlier this year, after the deal was first unveiled to the public. The NDRC did not release any additional details or expand on the specific justifications for blocking the acquisition in its public statement.

The proposed deal was unusual from its inception: it represented a rare instance of a large U.S. technology group acquiring an artificial intelligence firm with deep founding ties to China. Manus has gained recognition in global AI circles for developing a general-purpose AI agent capable of completing multi-step, complex work tasks without continuous human input. For Meta, the acquisition was expected to accelerate the company’s development of advanced AI capabilities to enhance product offerings across its entire ecosystem of social and digital platforms.

Even before the formal investigation was launched, Meta sought to address early regulatory concerns by confirming that after the acquisition closed, there would be no remaining Chinese ownership stakes in Manus, and the startup would end all its services and operations within mainland China. Still, Chinese regulatory bodies announced in January that they would launch a formal review to assess whether the deal aligned with the country’s existing laws and regulations governing foreign investment.

At the time of that announcement, China’s Ministry of Commerce emphasized that all companies engaging in outward investment, technology transfers, cross-border data flows, and cross-border acquisition transactions are required to comply fully with all relevant Chinese legal requirements. Meta has repeatedly noted that the vast majority of Manus’ employees are based in Singapore, the startup’s official registered headquarters.

In a written response to Monday’s ruling, Meta maintained that the proposed transaction had complied fully with all applicable laws. “We anticipate an appropriate resolution to the inquiry,” the California-headquartered company said in its statement, giving no further indication of what steps it might take moving forward. Industry analysts say the ruling underscores the increasing regulatory attention around the world to national security risks tied to AI and cross-border technology acquisitions, as global powers race to advance their own domestic artificial innovation ecosystems.

For decades, a world fully powered by autonomous, unmanned systems was nothing more than a far-fetched science fiction concept: aircraft that fly without human pilots, passenger vehicles that navigate roads without drivers, and delivery services that ship goods via flying drones. Today, that futuristic vision has moved from the pages of speculative stories to real-world demonstration, right in the heart of Luogang Park in Hefei, the capital city of East China’s Anhui province.

On April 24, two well-known international content creators, Oluwabunmi Jimoh and Joe Burns, traveled to the park to get a first-hand look at Hefei’s cutting-edge progress in the field of unmanned mobility. During their visit, the pair got up close to watch electric vertical take-off and landing (eVTOL) autonomous aircraft complete test operations, experienced a fully hands-free trip around the park on a driverless electric shuttle bus, and got a close look at the efficiency of the park’s automated drone delivery network, which can drop off orders directly to visitors across the green space.

The demonstration showcases how far China has advanced in developing and testing commercial-ready autonomous mobility solutions, positioning Hefei as a growing hub for innovation in the unmanned technology sector. Visitors and industry observers can access additional visual coverage of the event through a published original video to see the technology in action.

In a milestone display of aerospace cooperation between China and Pakistan, China has successfully placed the Pakistani remote sensing satellite PRSC-EO3 into its pre-planned orbit, completing a landmark launch mission from the Taiyuan Satellite Launch Center in northern China’s Shanxi Province.

The mission lifted off at 8:15 p.m. Beijing Time on Saturday, carried aloft by a Long March 6 carrier rocket. Following a standard ascent sequence, the satellite smoothly separated from the rocket and entered its designated orbit, marking a full success for the launch operation.

Beyond supporting Pakistan’s national space development goals, this mission also carries notable significance for China’s own Long March rocket program. It stands as the 640th flight mission completed across the entire Long March carrier rocket series, underscoring the consistent reliability and maturity of China’s domestic launch vehicle technology after decades of iterative development and real-world testing.

The launch follows a long track record of collaborative space projects between China and Pakistan, opening new opportunities for Pakistan to advance its capabilities in Earth observation, geographic surveying, and environmental monitoring through access to space infrastructure.

On April 24, a landmark demonstration of China’s emerging electric urban air mobility technology drew international attention, when Mexican digital creator Noelia Pascual took a three-minute autonomous test flight over Luogang Park in Hefei, the capital city of East China’s Anhui province. The flight was operated by local aerospace firm Hefei Heyi Aviation Co, using its electric vertical take-off and landing (eVTOL) aircraft, commonly referred to as an “air taxi”, which is currently undergoing trial testing for future commercial urban mobility applications.

Before boarding the autonomous aircraft, Pascual had prepared herself for a nerve-wracking, turbulent experience. But she left the flight far more impressed than she expected, praising the vehicle’s remarkable stability and robust safety design that eliminated most of her pre-flight anxiety.

“It’s actually very stable, very similar to the experience of flying in a helicopter, but it feels more stable than a helicopter,” Pascual shared enthusiastically with onlookers immediately after completing her flight.

She further noted that the transparent, real-time flight data displayed on an in-cabin screen played a key role in easing her concerns during the journey. “I never felt unsafe because the flight information is all displayed on the screen. You know what’s happening, and that gives you some reassurance,” she added.

The test flight comes as China continues to accelerate research and commercial development of eVTOL technology, positioning itself as a global leader in the emerging low-altitude economy sector. For international visitors and industry observers, the successful public demonstration offers a first-hand look at how close autonomous air mobility is to becoming a practical part of daily urban transportation.

On April 24, 2026, China marked another key milestone in its space-based internet infrastructure development with the successful launch of a batch of experimental satellites from the Xichang Satellite Launch Center located in the southwestern province of Sichuan. According to the China Aerospace Science and Technology Corporation, the country’s top state-owned space contractor, the mission lifted off at 2:35 p.m. Beijing Time, with a veteran Long March 2D liquid-fuel carrier rocket delivering the Space-based Internet Technology Demonstrator series satellites into their pre-planned orbits without incident.

This launch marks the ninth orbital deployment of satellites for the Space-based Internet Technology Demonstrator program, which kicked off with its inaugural mission back in July 2023. Among the new satellites placed into orbit is a platform developed by GalaxySpace, a leading private aerospace firm headquartered in Beijing. This particular spacecraft is designed to carry out cutting-edge technical trials for several critical next-generation satellite technologies, including broadband direct-to-device cellular communication, integrated space-ground network architecture, and other core enabling technologies for global satellite internet.

The launch is part of China’s broader, ambitious plan to build a large-scale low-Earth orbit satellite mega-constellation, which will consist of approximately 13,000 individual satellites working together to deliver comprehensive global internet coverage to users across the planet. This infrastructure will help bridge the digital divide for remote and underserved regions that lack access to traditional terrestrial broadband networks.

Produced by the Shanghai Academy of Spaceflight Technology, the Long March 2D rocket that carried out this mission is a proven workhorse of China’s launch fleet. Powered by liquid propellants, the rocket generates 300 metric tons of liftoff thrust, and is certified to deliver payloads of up to 1.2 tons into a 700-kilometer sun-synchronous orbit, making it well-suited for the deployment of this class of experimental communications satellites.

A groundbreaking new artificial intelligence model tailored specifically for the Tibetan language has cleared national regulatory approval and entered public pilot testing, marking a major milestone in expanding digital inclusion and technological development across China’s Tibetan-speaking regions. Developed by the State Key Laboratory of Tibetan Intelligence at Qinghai Normal University, the Zeta model — Qinghai’s first large-scale multimodal Tibetan language AI system — was officially unveiled in Beijing on April 22, 2026, opening new doors for innovation across sectors from cultural preservation to public services.

Unlike earlier Tibetan language AI tools that were limited to single functions such as basic text translation or speech recognition, Zeta was built from the ground up to deliver comprehensive, full-spectrum language processing capabilities across all major forms of linguistic interaction. According to Dorlha, executive deputy director of the development laboratory, the model supports integrated listening, speaking, reading, writing and translation across the three primary regional Tibetan dialects: Amdo, U-Tsang and Kham.

This broad capability set allows Zeta to tackle a wide range of specialized use cases that were out of reach for previous tools. Its core innovative functions include mixed-language document recognition, automated audiobook production, intelligent retrieval of ancient Tibetan literature, and real-time intelligent subtitle transcription. For industry-specific applications, the model also offers built-in features for digital broadcasting, agricultural information dissemination and tourist translation services, making it a flexible resource for public and private stakeholders across media, agriculture, tourism, healthcare, education and governance.

To address longstanding technical barriers in Tibetan language AI development — most notably the historical lack of large-scale, high-quality training data — the Zeta development team assembled an expansive, diverse training corpus. The model’s dataset includes 150 gigabytes of curated high-quality Tibetan text, 87 million parallel multilingual sentence pairs across Tibetan, standard Chinese and English, and 30,000 hours of labeled multi-dialect Tibetan audio recordings. Zeta integrates all three languages into a unified multilingual framework, and pairs custom-developed algorithms with full compatibility for domestic AI infrastructure, delivering proven technical maturity and room for future expansion. It is available in three parameter configurations of 7 billion, 50 billion and 122 billion parameters to accommodate different use cases and computing environments, from mobile device deployment to large-scale server-side applications.

Nyima Tashi, director of the State Key Laboratory of Tibetan Intelligence and a professor at Xizang University, emphasized that the launch of Zeta and its supporting applications will drive high-quality economic and social development across China’s Tibetan regions. Moving forward, the research team plans to continue expanding the model’s capabilities by opening its multimodal functions through public application programming interfaces, fostering deeper collaboration between academic institutions and private sector enterprises, and building a complete, self-sustaining ecosystem for Tibetan language AI innovation. The lab also plans to increase research investment, strengthen specialized talent training, and advance partnerships across industry, academia and research institutions to further refine the technology.

Zeta’s launch comes just one month after the release of Deep-Zang, the first large Tibetan language model developed in the Xizang Autonomous Region, giving users across Tibetan-speaking regions a growing range of specialized AI tools to meet their needs. For Tibetan communities and users, the innovation carries far more meaning than just technological progress. Tenzin Palden, a Tibetan student studying at Shandong Agricultural University, noted that Zeta addresses long unmet needs for advanced Tibetan language digital tools, offering new hope for preserving Tibetan linguistic and cultural identity in an increasingly digital-first world.

“By addressing historical challenges like limited datasets and diversity in Tibetan dialects, this innovation provides much-needed momentum for bridging the wisdom of Tibetan traditions with modern development,” Tenzin Palden said. “It is not just a technological achievement but also a reflection of the protection and transmission of ethnic culture.”

As competition in artificial intelligence between the United States and China reaches new levels of intensity, prominent Chinese AI startup DeepSeek rolled out previews of its highly anticipated next-generation V4 model lineup on Friday, marking another major milestone in China’s push to advance its domestic AI ecosystem independent of U.S. technology.

The V4 release comes months after DeepSeek’s specialized R1 reasoning model upended global tech markets earlier this year, with the startup claiming it outperformed comparable U.S.-built models at a far lower cost. R1 quickly became a global symbol of China’s rapid progress closing the AI gap with the United States, and expectations for V4 have been building among developers and users eager to compare its capabilities to leading models from U.S. industry leaders OpenAI, Anthropic, and Google. Some analysts initially projected the V4 launch would arrive more than two months ago, to coincide with the Lunar New Year holiday.

DeepSeek’s new V4 family includes two core open-source variants: the high-performance V4 Pro and the lightweight V4 Flash. The startup says the new models deliver sweeping upgrades across three key areas: general knowledge retention, logical reasoning, and agentic functionality — the ability for AI to complete complex, multi-step workflows and tasks without constant human input. One of the most notable shifts in the new lineup is its underlying hardware: unlike prior DeepSeek models that relied on U.S.-made chips from industry leader Nvidia, V4 is powered by chips developed by Chinese tech giant Huawei.

In a statement accompanying the launch, DeepSeek shared internal benchmark results comparing V4 to top U.S. models. The company notes its top-tier V4 Pro Max delivers superior performance on standard reasoning tests compared to OpenAI’s recently released GPT-5.2 and Google’s Gemini 3.0-Pro, though it falls slightly short of OpenAI’s GPT-5.4 and Google’s Gemini 3.1-Pro. The V4 Pro, DeepSeek claims, outperforms Anthropic’s mid-tier Claude Sonnet 4.5 in agentic capabilities and comes close to matching Anthropic’s flagship Claude Opus 4.5. For everyday simple agent tasks, the more efficient V4 Flash matches the performance of V4 Pro, with reasoning capabilities that nearly equal its higher-end counterpart, per the company’s testing.

The V4 launch came just hours after OpenAI introduced its own newest model, GPT-5.5, in a clear sign of the breakneck pace of competition in the global AI race. Both V4 variants also include a game-changing 1 million token context window — the measure of how much text and data an AI can process and retain in a single session. That marks an eightfold increase from the 128,000 token window supported by DeepSeek’s previous V3 model, released in late 2024, and enables the new models to handle far larger datasets, long documents, and complex extended conversations more effectively. The startup also emphasized the V4 lineup is designed to run far more efficiently than prior generations.

Unlike closed, proprietary models from leading U.S. AI developers, DeepSeek makes its core technology open source, allowing outside developers to modify, adapt, and build new tools on top of its models. The company also offers a free publicly accessible chatbot for web and mobile users, helping it gain a large global user base. A January report from Microsoft found DeepSeek usage has grown rapidly across many developing nations, especially in regions where Huawei smartphones dominate the consumer market.

Huawei confirmed its compatibility with the new models in a separate statement Friday, noting its Ascend chip platform and supporting ecosystem work seamlessly with DeepSeek V4. Marina Zhang, an associate professor at the University of Technology Sydney, described the V4 rollout as a “pivotal milestone for China’s AI industry”, particularly amid rising global pressure for countries to build self-reliance in critical emerging technologies. She added that the partnership with Huawei demonstrates that a fully functional Chinese AI ecosystem independent of Nvidia’s market dominance is technically achievable, even as U.S.-China technological decoupling continues.

Lian Jye Su, chief analyst at global technology research firm Omdia, concluded that “Based on the benchmark results, it does appear DeepSeek V4 is going to be very competitive against its U.S. rivals.”

Not all industry observers are convinced the V4 represents a transformative leap forward, however. Ivan Su, senior equity analyst at investment research firm Morningstar, argued that while V4 is a solid, capable update to DeepSeek’s product lineup, it does not deliver the same level of groundbreaking innovation that R1 introduced earlier this year. He noted that domestic competition within China’s AI sector has intensified dramatically since R1’s launch, and that independent third-party testing is needed to verify DeepSeek’s own performance claims, since the startup’s internal comparisons cannot yet be confirmed by outside experts.

The V4 launch comes amid ongoing friction between U.S. AI firms and Chinese developers over intellectual property. Earlier this year, Anthropic publicly accused DeepSeek and two other China-based AI labs of running “industrial-scale campaigns” to steal its technology via a process called knowledge distillation, a method that trains a smaller model by feeding it the outputs of a more powerful competitor model to replicate its capabilities. OpenAI made similar accusations in a letter to U.S. lawmakers, and this week Michael Kratsios, chief science and technology adviser to U.S. President Donald Trump, repeated the claims, accusing Chinese tech firms of distilling leading U.S. AI systems to “exploit American expertise and innovation.”

The Chinese embassy in Washington has pushed back against these allegations, framing them as unjustified efforts by the United States to stifle competition from Chinese tech companies.