Connect with us

Published

on

China is focusing on large language models (LLMs) in the artificial intelligence space. 

Blackdovfx | Istock | Getty Images

China is embracing open-source AI models in a trend market watchers and insiders say is boosting AI adoption and innovation in the country, with some suggesting it is an ‘Android moment’ for the sector.

The open-source shift has been spearheaded by AI startup DeepSeek, whose R1 model released earlier this year challenged American tech dominance and raised questions over Big Tech’s massive spending on large language models and data centers. 

While R1 created a splash in the sector due to its performance and claims of lower costs, some analysts say the most significant impact of DeepSeek has been in catalyzing the adoption of open-source AI models. 

“DeepSeek’s success proves that open-source strategies can lead to faster innovation and broad adoption,” said Wei Sun, principal analyst of artificial intelligence at Counterpoint Research, noting a large number of firms have implemented the model. 

“Now, we see that R1 is actively reshaping China’s AI landscape, with large companies like Baidu moving to open source their own LLMs in a strategic response,” she added. 

On March 16, Baidu released the latest version of its AI model, Ernie 4.5, as well as a new reasoning model, Ernie X1, making them free for individual users. Baidu also plans to make the Ernie 4.5 model series open-source from end-June. 

Experts say that Baidu’s open-source plans represent a broader shift in China, away from a business strategy that focuses on proprietary licensing. 

“Baidu has always been very supportive of its proprietary business model and was vocal against open-source, but disruptors like DeepSeek have proven that open-source models can be as competitive and reliable as proprietary ones,” Lian Jye Su, chief analyst with technology research and advisory group Omdia previously told CNBC.

Open-source vs proprietary models

Open-source generally refers to software in which the source code is made freely available on the web for possible modification and redistribution.

AI models that call themselves open-source had existed before the emergence of DeepSeek, with Meta‘s Llama and Google‘s Gemma being prime examples in the U.S. However, some experts argue that these models aren’t really open source as their licenses restrict certain uses and modifications, and their training data sets aren’t public.

DeepSeek’s R1 is distributed under an ‘MIT License,’ which Counterpoint’s Sun describes as one of the most permissive and widely adopted open-source licenses, facilitating unrestricted use, modification and distribution, including for commercial purposes.

The DeepSeek team even held an “Open-Source Week” last month, which saw it release more technical details about the development of its R1 model. 

While DeepSeek’s model itself is free, the start-up charges for Application Programming Interface, which enables the integration of AI models and their capabilities into other companies’ applications. However, its API charges are advertised to be far cheaper compared with OpenAI and Anthropic’s latest offerings.

OpenAI and Anthropic also generate revenue by charging individual users and enterprises to access some of their models. These models are considered to be ‘closed-source,’ as their datasets, and algorithms are not open for public access.

China opens up

In addition to Baidu, other Chinese tech giants such as Alibaba Group and Tencent have increasingly been providing their AI offerings for free and are making more models open source.

For example, Alibaba Cloud said last month it was open-sourcing its AI models for video generation, while Tencent reportedly released five new open-source models earlier this month with the ability to convert text and images into 3D visuals.

Smaller players are also furthering the trend. ManusAI, a Chinese AI firm that recently unveiled an AI agent that claims to outperform OpenAI’s Deep Research, has said it would shift towards open source.

“This wouldn’t be possible without the amazing open-source community, which is why we’re committed to giving back” co-founder Ji Yichao said in a product demo video. “ManusAI operates as a multi-agent system powered by several distinct models, so later this year, we’re going to open source some of these models,” he added.

Zhipu AI, one of the country’s leading AI startups, this month announced on WeChat that 2025 would be “the year of open source.”

Ray Wang, principal analyst and founder of Constellation Research, told CNBC that companies have been compelled to make these moves following the emergence of DeepSeek.

“With DeepSeek free, it’s impossible for any other Chinese competitors to charge for the same thing. They have to move to open-source business models in order to compete,” said Wang. 

AI scholar and entrepreneur Kai-Fu Lee also believes this dynamic will impact OpenAI, noting in a recent social media post that it would be difficult for the company to justify its pricing when the competition is “free and formidable.”

“The biggest revelation from DeepSeek is that open-source has won,” said Lee, whose Chinese startup 01.AI has built an LLM platform for enterprises seeking to use DeepSeek.

U.S.-China competition

OpenAI — which started the AI frenzy when it released its ChatGPT bot in November 2022— has not signaled that it plans to shift from its proprietary business model. The company which started as a nonprofit in 2015 is moving towards towards a for-profit structure.

Sun says that OpenAI and DeepSeek both represent very different ends of the AI space. She adds that the sector could continue to see division between open-source players that innovate off one another and closed-source companies that have come under pressure to maintain high-cost cutting-edge models. 

The open-source trend has put in to question the massive funds raised by companies such as OpenAI. Microsoft has invested $13 billion into the company. It is in talks to raise up to $40 billion in a funding round that would lift its valuation to as high as $340 billion, CNBC confirmed at the end of January.

In September, CNBC confirmed the company expects about $5 billion in losses, with revenue pegged at $3.7 billion revenue. OpenAI CFO Sarah Friar, has also said that $11 billion in revenue is “definitely in the realm of possibility” for the company this year.

China's open-source AI push is an Android moment and a huge sentiment boost: Hedge fund manager

On the other hand, Chinese companies have chosen the open-source route as they compete with the more proprietary approach of U.S. firms, said Constellation Research’s Wang. “They are hoping for faster adoption than the closed models of the U.S.,” he added. 

Speaking to CNBC’s “Street Signs Asia” on Wednesday, Tim Wang, managing partner of tech-focused hedge fund Monolith Management, said that models from companies such as DeepSeek have been “great enablers and multipliers in China,” demonstrating how things can be done with more limited resources.

According to Wang, open-source models have pushed down costs, opening doors for product innovation — something he says Chinese companies historically have been very good at.

He calls the development the “Android moment,” referring to when Google’s Android made its operating system source code freely available, fostering innovation and development in the non-Apple app ecosystem.

“We used to think China was 12 to 24 months behind [the U.S.] in AI and now we think that’s probably three to six months,” said Wang.

However, other experts have downplayed the idea that open-source AI should be seen through the lens of China and U.S. competition. In fact, several U.S. companies have integrated and benefited from DeepSeek’s R1. 

“I think the so-called DeepSeek moment is not about whether China has better AI than the U.S. or vice versa. It’s really about the power of open-source,” Alibaba Group Chairperson Joe Tsai told CNBC’s CONVERGE conference in Singapore earlier this month. 

Tsai added that open-source models give the power of AI to everyone from small entrepreneurs to large corporations, which will lead to more development, innovation and a proliferation of AI applications.

Continue Reading

Technology

U.S. blacklists over 50 Chinese companies in bid to curb Beijing’s AI, chip capabilities

Published

on

By

U.S. blacklists over 50 Chinese companies in bid to curb Beijing's AI, chip capabilities

William_potter | Istock | Getty Images

The U.S. on Tuesday added dozens of Chinese tech companies to its export blacklist in its first such effort under the Donald Trump administration, as it doubles down on curtailing Beijing’s artificial intelligence and advanced computing capabilities.

The U.S. Department of Commerce’s Bureau of Industry and Security added 80 organizations to an “entity list,” with more than 50 from China, barring American companies from supplying to those on the list without government permits.

The companies were blacklisted for allegedly acting contrary to U.S. national security and foreign policy interests, the agency said, as part of its efforts to further restrict Beijing’s access to exascale computing tech, which can process vast amounts of data at very high speeds, as well as quantum technologies.

Dozens of Chinese entities were targeted for their alleged involvement in developing advanced AI, supercomputers and high-performance AI chips for military purposes, the Commerce Department said, adding that two firms were supplying to sanctioned entities such as Huawei and its affiliated chipmaker HiSilicon.

It blacklisted 27 Chinese entities for acquiring U.S.-origin items to support China’s military modernization and seven firms for helping advance China’s quantum technology capabilities.

Among the organizations in the “entity list” were also six subsidiaries of Chinese cloud-computing firm Inspur Group, which had been blacklisted by the Joe Biden administration in 2023.

The latest additions “cast an ever-widening net aimed at third countries, transit points and intermediaries,” said Alex Capri, a senior lecturer at National University of Singapore and author of “Techno-Nationalism: How it’s reshaping trade, geopolitics and society.”

Chinese firms have managed to gain access to U.S. strategic dual-use technologies via certain third parties, he said, referring to loopholes that have allowed Chinese companies access to U.S. technologies despite restrictions.

“U.S. officials will continue to step up tracking and tracing operations aimed at the smuggling of advanced semiconductors made by Nvidia and Advanced Micro Devices,” he said.

The expanded export restrictions come at a time when tensions between Washington and Beijing have been rising with the Trump administration ratcheting up tariffs against China.

The rapid rise of Chinese AI startup DeepSeek has boosted the adoption of open-source low-cost AI models in China, putting pressure on leading U.S. competitors with higher-cost, proprietary models.

The chip industry wants more clarity around policymaking, says 'Chip War' author Chris Miller

The Biden administration imposed sweeping export controls against China, encompassing everything from semiconductors to supercomputers under the so-called “small yard, high fence” policy. The approach aims to place restrictions on a small number of technologies with significant military potential while maintaining normal economic exchange in other areas.

Under Secretary of Commerce for Industry and Security Jeffrey I. Kessler said the agency was “sending a clear, resounding message” that the Trump administration will prevent U.S. technologies from “being misused for high performance computing, hypersonic missiles, military aircraft training, and UAVs (unmanned aerial vehicle) that threaten our national security.”

“The entity list is one of many powerful tools at our disposal to identify and cut off foreign adversaries seeking to exploit American technology for malign purposes,” he added.

Inspur Group and Huawei did not immediately respond to CNBC’s requests for comment.

Continue Reading

Technology

Amazon is testing shopping, health assistants as it pushes deeper into generative AI

Published

on

By

Amazon is testing shopping, health assistants as it pushes deeper into generative AI

Amazon CEO Andy Jassy speaks during an Amazon Devices launch event in New York City, U.S., February 26, 2025. 

Brendan Mcdermid | Reuters

Amazon, in an effort to infuse generative artificial intelligence across a wider swath of its e-commerce universe, recently began testing a shopping assistant and a health-focused chatbot with a subset of users.

AI has become a major area of investment across Amazon, including in its retail, cloud computing, devices and health-care businesses. Within the retail business, Amazon has already launched a shopping chatbot, an AI assistant for sellers and AI shopping guides.

The new services Amazon is testing appeared on its app or website in recent weeks. The shopping tool, called Interests AI, prompts users to describe an interest “using your own words,” and then it generates a curated selection of products. The feature lets consumers browse for products using more conversational language and is separate from the main search bar on Amazon’s website.

Amazon’s Interests AI feature lets users input more conversational search queries

Amazon

Within its core app, Amazon has a landing page for the feature.

“Describe your interest, like ‘coffee brewing gadgets’ or ‘latest pickleball accessories’ — and we’ll find relevant products for you,” the page says. Other suggested searches include “children books about persistence and dealing with failure,” and “brain teasers that are not too hard, made out of wood or metal.”

Amazon CEO Andy Jassy said last month that employees have built or are in the process of building roughly 1,000 generative AI applications across the company. Its cloud unit offers a chatbot for businesses, called Q. In commerce, the company has rolled out services for consumers as well as its millions of third-party sellers.

Amazon is also exploring ways that AI can address medical needs. The company is testing a chatbot on its website and mobile app called “Health AI,” which can answer health and wellness questions, “provide common care options for health care needs,” and suggest products.

While Rufus, Amazon’s shopping chatbot, can suggest products like ice packs and ibuprofen, Health AI goes further, providing users with medical guidance and care tips, such as how to deal with cold symptoms or the flu. The site says the service can’t provide personalized medical advice.

Some responses feature a “clinically verified” badge, which denotes information that’s been “reviewed by US-based licensed clinicians,” Amazon says.

Health AI also steers users to Amazon’s online pharmacy, along with clinical services offered by One Medical, the primary care provider it acquired for roughly $3.9 billion in 2022.

Amazon recently began testing a health-related AI assistant that can provide medical guidance and suggest products.

Amazon

More consumers are embracing generative AI as a shopping tool, and with features like Health AI and Interests AI, Amazon wants shoppers to use its own services over rivals like OpenAI’s ChatGPT.

With enough use, Amazon could gain valuable insights on the ways that people are interacting with AI assistants as the company prepares to overhaul Alexa, the digital assistant it launched more than a decade ago.

Amazon announced Alexa+, a new version of the technology embedded with generative AI, late last month. The company says that Alexa+, which has yet to roll out, is capable of handling more complex tasks and can serve as an “agent” by taking actions for users without their direct involvement.

Andrew Bell, an Amazon e-commerce manager for the National Fire Protection Association who also publishes research on Amazon’s patent filings and AI development, came across the new shopping and health features and recently posted about them on LinkedIn.

Bell said in an interview that Alexa+ could potentially draw upon models developed for Amazon applications like Health AI to answer queries.

“If there’s a health-related question, Alexa+ is going to maybe call on Health AI,” Bell said. “If there’s a product-related question, Alexa+ can call on Rufus.”

Amazon didn’t provide a comment for this story.

WATCH: Amazon’s SVP of Devices on Alexa+

Alexa+ has arrived: Amazon SVP Devices on the new AI-powered assistant

Continue Reading

Technology

With 23andMe entering bankruptcy, here’s how to delete your genetic data

Published

on

By

With 23andMe entering bankruptcy, here's how to delete your genetic data

Signage at 23andMe headquarters in Sunnyvale, California, U.S., on Wednesday, Jan. 27, 2021.

David Paul Morris | Bloomberg | Getty Images

23andMe has officially filed for Chapter 11 bankruptcy protection, which means its assets — including its vast genetic database — will soon be up for sale.

The company continues to sell its at-home DNA testing kits, allowing consumers to get insight into their family histories and genetic profiles. DNA data is particularly sensitive because each person’s sequence is unique, meaning it can never be fully anonymized, according to the National Human Genome Research Institute.

If genetic data falls into the hands of bad actors, it could be used to facilitate identity theft, insurance fraud or other crimes. 23andMe has been plagued by privacy concerns in recent years after hackers accessed the information of nearly 7 million customers in October 2023.

As part of the bankruptcy process, the company said it will seek a partner that shares its commitment to customer data privacy, and that there will be no changes to how it stores, manages and protects data through the sale process.

“Our users’ privacy and data are important considerations in any transaction, and we remain committed to our users’ privacy and to being transparent with our customers about how their data is managed,” the company said in an FAQ page about the bankruptcy filing. “Any buyer of 23andMe will be required to comply with applicable law with respect to the treatment of customer data.”

Still, experts and officials are urging 23andMe customers to proceed with caution. California Attorney General Rob Bonta on Friday issued a consumer alert, encouraging residents to consider deleting their genetic data from 23andMe, which is based in his home state.

“Given 23andMe’s reported financial distress, I remind Californians to consider invoking their rights and directing 23andMe to delete their data and destroy any samples of genetic material held by the company,” Bonta said in the release.

Adrianus Warmenhoven, who serves on the security advisory board at NordVPN, described genetic data as the blueprint of your entire biological profile.” He encouraged consumers to delete their information and be mindful of the companies they chose to share it with going forward.

“Monitor your digital footprint regularly, and you can also sign up for credit monitoring or identity theft protection services,” Warmenhoven said in a statement to CNBC. “Revoke permissions you no longer require, shut down any account you don’t use, and learn about how your data is used.” 

23andMe said customers can still delete their account and accompanying data. Here’s how:

Delete your genetic data from 23andMe

  • Go to 23andMe.com and sign in to your account.
  • Click on your profile in the upper righthand corner of the site, then click “Settings.”
  • Scroll to the section at the very bottom of the page called “23andMe Data” and click the oval button that says “View.”
  • Check the boxes of any data you would like to download and click “Request Download.” This step is optional and can take up to 30 days. You can continue with the following steps while you wait.
  • Scroll to the bottom of the page and click the red button that says “Permanently Delete Data.”
  • You will receive an email with the subject line “23andMe Delete Account Request.” Open it, and click the button that says “Permanently Delete All Records.” Your data will not be deleted unless you complete this step.

At this point, your personal information and your account will be permanently deleted from 23andMe, according to the deletion email from the company. Additionally, your data will not be used in any future research projects, and any personal samples the company was storing will be discarded.

WATCH: The rise and fall of 23andMe

The rise and fall of 23andMe

Continue Reading

Trending