China’s artificial intelligence race has entered a new phase, with Beijing-based startup Z.ai emerging as the latest company challenging the dominance of leading U.S. AI developers. Its recently launched GLM-5.2 model is winning praise from developers, technology executives and investors, bolstering the belief that China’s AI ecosystem is closing the performance gap with OpenAI and Anthropic while maintaining a significant cost advantage.

The model, launched last month, is generating growing interest across the global developer community because of its advanced coding and autonomous agent capabilities, allowing it to complete sophisticated software engineering and reasoning tasks with minimal human prompting.

According to a Reuters report, industry observers describe the enthusiasm surrounding GLM-5.2 as a “mini DeepSeek moment,” recalling the shockwaves created when DeepSeek unveiled a powerful low-cost reasoning model early last year that challenged assumptions about the enormous capital required to build frontier AI.

Unlike earlier generations of Chinese AI models, which were often viewed as cheaper but less capable alternatives to U.S. offerings, GLM-5.2 is increasingly being discussed as a genuine competitor to the latest systems from OpenAI and Anthropic.

Its rapid adoption is evident on OpenRouter, one of the world’s leading AI developer platforms, where GLM-5.2 has climbed above Anthropic’s models in usage rankings. The model has also received endorsements from influential technology leaders, including Snowflake CEO Sridhar Ramaswamy and venture capitalist Marc Andreessen, further boosting its credibility among software developers.

David Sacks, who previously served as U.S. President Donald Trump’s AI czar, said the emergence of GLM-5.2 demonstrates how rapidly China’s AI capabilities are advancing.

“We now have a Chinese open-weight model that is as good as the currently available models from OpenAI and Anthropic,” Sacks said last week, before Washington lifted restrictions on Anthropic’s Fable and Mythos models on Tuesday.

Speaking on the All-In podcast, Sacks added that GLM-5.2 is “just a tick below Opus 4.8 (from Anthropic) and right up there with GPT 5.5 (from OpenAI),” warning that “we cannot afford to do things that slow our companies down.”

Within parts of the U.S. technology industry, there is growing concern that regulatory uncertainty could weaken America’s lead in artificial intelligence just as Chinese companies are becoming more competitive.

Several analysts believe the timing has also favored Z.ai.

Washington’s temporary restrictions on Anthropic’s newest models and OpenAI’s delayed public rollout of GPT-5.6 have prompted many developers to experiment with alternative models, accelerating international interest in GLM-5.2.

Brian Tse, founder and CEO of Beijing-based AI consultancy Concordia AI, said developers are increasingly seeking alternatives to proprietary American models.

“The international developer community is increasingly aware that relying solely on proprietary, U.S.-based API models carries significant risk,” Tse said.

Cost has become another powerful advantage.

As businesses deploy increasingly sophisticated AI agents, token consumption—the units used to measure AI usage—has risen sharply, making proprietary AI services substantially more expensive.

Against that backdrop, GLM-5.2 has attracted attention by delivering performance approaching frontier models while costing roughly one-sixth as much as leading closed-source offerings from OpenAI and Anthropic. Although Z.ai has not disclosed how much it spent developing GLM-5.2, the pricing has made it particularly attractive to startups, software developers and enterprises looking to control AI infrastructure costs without sacrificing capability.

Independent benchmarks reinforce its growing reputation.

GLM-5.2 currently ranks fifth on Artificial Analysis’ large language model intelligence leaderboard, which measures overall reasoning, knowledge, and coding capabilities across numerous standardized tests. It also ranks second on Code Arena’s front-end coding leaderboard, which evaluates models’ ability to generate websites and user interfaces.

For many developers, however, the biggest attraction lies in usability rather than benchmark scores.

Tiezhen Wang, former Asia-Pacific lead at Hugging Face, said GLM-5.2 significantly lowers the technical barriers traditionally associated with deploying open-source AI.

“The shift GLM-5.2 brings is that the open-source model has become a plug-and-play, out-of-the-box product,” Wang said.

“You just deploy the model and without doing any complex fine-tuning systems, it is in a highly usable, ready-to-use state. This drastically lowers the barrier to entry for open-source adoption.”

Z.ai’s ambitions extend well beyond its current model.

In a response to Elon Musk on X last month, founder Tang Jie said the company aims to produce an AI model comparable to Anthropic’s Fable before the end of the first quarter next year, signaling its intention to compete directly with the world’s most advanced AI systems.

Even so, major challenges remain before GLM-5.2 can achieve widespread enterprise adoption outside China. Data security and geopolitical concerns continue to discourage many Western corporations, particularly banks, government agencies and cybersecurity firms, from incorporating Chinese AI models into critical systems.

Wei Sun, principal AI analyst at Counterpoint Research, said regulatory concerns remain a significant obstacle.

“I have seen some discussion among European companies about whether it could be used in enterprise settings,” Sun said.

“In the EU and U.S., some clients, partners and regulated industries may simply be unwilling to accept Chinese models in their AI stack, regardless of technical performance or price.”

Enterprise adoption also tends to move slowly because replacing AI infrastructure often requires months of testing, integration and regulatory review.

Nevertheless, some analysts argue that those concerns may be less significant than many assume. They note that companies can deploy open-weight Chinese models on their own servers or through U.S.-based cloud providers, limiting data exposure while benefiting from lower costs and greater flexibility.

Poe Zhao, founder of the Hello China Tech newsletter, said practical considerations often outweigh geopolitical ones among developers.

“Developers tend to care less about where a model comes from than whether it works, how much it costs and whether they can deploy or access it reliably,” Zhao said.

“The likely pattern is partial routing, not overnight replacement of OpenAI or Anthropic. So yes, it is a mini DeepSeek moment but in a narrower, developer-centric sense.”

Evidence suggests Chinese AI models have already been gaining international traction since DeepSeek disrupted the industry. A report published earlier this year by RAND found that Chinese large language models increased their global market share from 3% to 13% during the two months following DeepSeek’s R1 launch. The gains were particularly pronounced across developing economies and countries maintaining close economic and political ties with Beijing.

The release of DeepSeek’s low-cost reasoning model also triggered a global technology selloff by challenging the assumption that only companies spending hundreds of billions of dollars on AI infrastructure could compete at the frontier.

GLM-5.2 now appears to be extending that narrative. Rather than simply offering a low-cost alternative, Z.ai is demonstrating that Chinese AI developers are capable of producing models that approach the performance of leading American systems while remaining substantially cheaper to deploy. Although regulatory barriers and trust issues are likely to slow adoption among large Western enterprises, the model’s rapid acceptance among developers indicates that China’s AI ecosystem is becoming a more formidable competitor in the global race for artificial intelligence leadership.

India’s aggressive efforts to combat the proliferation of fake websites impersonating major brands are raising alarms among global domain registrars, who warn that new court-mandated measures could undermine internet privacy and create unintended consequences for legitimate businesses worldwide.

Per Reuters, the world’s biggest internet domain seller, GoDaddy, has mounted a strong legal challenge against directives issued by a New Delhi court that it says rewrite the rules of internet governance, potentially exposing users to greater risks while attempting to address a serious problem of online deception.

Soaring smartphone and internet use has coincided with a worsening problem of online fraud in India, the world’s most populous nation. It is a key challenge for Prime Minister Narendra Modi’s government, which last year received 2.4 million complaints of alleged cyber fraud amounting to $2.4 billion.

Starting in 2019, lawsuits were brought by dozens of Indian and global firms — Amazon against fake shopping sites trading on its name and McDonald’s complaining against bogus sites offering franchises. In December, an Indian court blocked more than 1,100 such websites.

The New Delhi judge, however, went further, ordering sweeping new measures that tech experts say have rewritten rules of internet governance: domain sellers should not offer buyers free privacy protection by default, the buyer’s details should be released to anyone with a “legitimate interest” within 72 hours, and website addresses that are variations of protected brand names must be prohibited.

U.S.-based GoDaddy has challenged the directives before a larger bench of judges at the Delhi High Court. It argues the ruling will affect legitimate businesses that have names similar to big brands.

Stopping privacy-by-default features, GoDaddy said, will result in public disclosure of the names, addresses, phone numbers, and emails of legitimate website owners, exposing them to “foreseeable privacy and security risks” such as stalking and harassment.

As domain names operate globally, not locally, the order could force GoDaddy to regulate website addresses across the world, it said.

On the court’s order imposing a 72-hour deadline on companies to provide registration details to anyone with “legitimate interest,” GoDaddy argues it has no wherewithal to assess who has a legitimate interest or not.

The “commercially destabilizing” directives may force domain name companies to “exit India,” said one of GoDaddy’s appeal documents that ran into 5,121 pages.

“Engines for large scale deception,” the December ruling noted about the fake websites.

One of the 14 measures outlined by the court said masking a domain buyer’s registration details should now be offered as a payable service, as the feature acts “as a cloak” to hide the identity of rogue operators.

Despite the court order, which remains in force, GoDaddy’s website still promotes its offering as one that includes “free privacy protection forever… we redact your name, address, phone number and email” from the public directory.

GoDaddy argues that diluting the privacy feature will run contrary to India’s data protection law and the European Union’s GDPR law, which mandates a “privacy by default” approach.

Farzaneh Badii, a New York-based researcher on internet governance, criticized the New Delhi ruling, noting that Europe redacted such details because publishing them had been abused by harassment and targeted phishing.

“The people exposed will be journalists, activists, small business owners, and private individuals. The brand impersonators will not,” she said.

In cases like the one brought by McDonald’s, the company sought action against 110 websites like mcdonaldsfranchiseindia.com, with some using its Golden Arches logo and selling fake franchises for “huge sums of money.”

After blocking those, GoDaddy says the court’s additional bar on offering alphanumeric variations of a trademark once it is protected, like McDonald’s, will act like “blanket injunctions,” which are difficult to implement.

The word “McDonald” is of Scottish origin and derived from a name meaning “son of the world ruler,” GoDaddy said, adding that an injunction against using it will effectively “confer a monopoly” over a common name with linguistic and historical meaning.

Reuters found domains like mcdonalds-india-franchise.com were still available on GoDaddy India for around $10. The U.S. giant also submitted research compiled from Merriam-Webster’s website to argue that safeguarding variations of a protected trademark like “HUL”, Unilever’s Indian unit, could overlap with 118 English words that contain the string, like “hulk” and “moghul.”

It is “virtually impossible to register a domain name containing an English word that does not overlap with a registered trademark,” GoDaddy says.

Government and Industry Perspectives

Modi’s Home Minister, Amit Shah, said this year that one person falls prey to cybercrime every 37 seconds in India, and a lack of action risks turning the menace into a “national crisis.”

While the sweeping December directives were issued by a court, they followed government submissions, documents showed.

An unreported 59-page IT ministry document from 2023, contained in GoDaddy’s latest appeal papers, revealed New Delhi conveyed to the judge it was concerned about the “issue of domain name abuse” and “lack of stringent verification.”

The home ministry, tasked with handling cybercrime, told the judge registration details “should be readily (made) available” for investigations.

That stand is in line with Modi’s bitter disagreements and spats with global technology giants in recent years. New Delhi has repeatedly criticized companies like Meta, X, Google, and Telegram, and even taken some of them to court, for not doing enough to police content it sees as against national interests.

The judges will hear the appeals on July 16.

Global Ramifications of Local Rules

With an annual revenue of $5 billion, GoDaddy manages 80 million domains and serves over 20 million users. In 2024, company executives said India was its biggest region in the emerging market space.

GoDaddy rivals, Arizona-based Namecheap and Netherlands-based Hosting Concepts, have also challenged the New Delhi ruling, court records show.

The legal dispute embroiling GoDaddy and others was triggered by more than 20 companies that sought the court’s intervention against fake websites damaging their brand. These included Amazon, McDonald’s, Microsoft, Xiaomi, and Colgate-Palmolive.

Deborah O’Neill, a senator from the ruling Labor Party who made the whistleblower allegations against KPMG public in March, said the “unethical” culture exposed by successive scandals needed to be disrupted. She warned the firms would “fight tooth and nail” to keep their existing structures, “because it is in their financial interest to do so”.

Barbara Pocock, a Greens senator who has campaigned for tougher regulation of the sector, said the government already knew what the solutions were from its previous inquiries, and called for urgent action.

“Labor needs to put an end to the Big Four’s special treatment and regulate them like other Australian businesses,” she said in a statement.

Nvidia is broadening its role in the artificial intelligence industry beyond designing the chips powering the AI revolution, unveiling a new revenue-sharing initiative that will allow fast-growing startups to exchange future earnings for access to high-performance computing infrastructure.

The strategy stemmed from the company’s ambition to become not just the dominant supplier of AI hardware but also a long-term financial partner to the next generation of AI companies.

The initiative, announced on Thursday, comes as access to computing power has become one of the biggest constraints on AI development globally. Training and deploying advanced AI models requires enormous numbers of graphics processing units (GPUs), most of which are supplied by Nvidia. That scarcity has created an environment where compute capacity has become as valuable as capital, prompting startups to seek alternative financing arrangements that reduce upfront infrastructure costs.

Under the new partnership program, Nvidia will provide eligible AI startups with token credits that grant access to computing infrastructure powered by its chips. Instead of paying entirely in cash, participating companies will share a portion of future product revenue and cloud-services income with Nvidia, creating a financing model that aligns the chipmaker’s returns with the commercial success of its customers.

The arrangement effectively transforms Nvidia from a hardware supplier into an infrastructure financier, enabling promising AI companies to scale faster while giving Nvidia exposure to the future revenues of businesses built on its technology.

The company said the program is designed for cloud-native AI companies, foundation model developers and enterprise AI firms. Nvidia will act as an intermediary, helping startups secure full-stack computing infrastructure that combines its industry-leading GPUs with networking, software and cloud resources.

Building An AI Ecosystem, Not Just Selling Chips

The latest initiative underpins how Nvidia is steadily expanding its influence across every layer of the AI value chain. For years, the company generated most of its revenue by selling GPUs to hyperscale cloud providers such as Amazon Web Services, Microsoft Azure, Google Cloud, and Oracle Cloud. Those companies, in turn, rented computing capacity to AI developers.

The new model allows Nvidia to participate more directly in the growth of AI startups themselves, giving it exposure to recurring revenue rather than relying solely on hardware sales.

Industry analysts have described Nvidia as evolving into an AI infrastructure platform rather than simply a semiconductor manufacturer. The company now offers chips, networking equipment, software frameworks such as CUDA, AI development platforms, cloud partnerships and, increasingly, financial structures that help customers gain access to computing resources.

The strategy also strengthens customer loyalty by tying startups more closely to Nvidia’s technology stack from the earliest stages of development.

Nvidia named two Australian companies as the initial infrastructure partners supporting the program. Sharon AI plans to deploy up to 40,000 Nvidia GPUs, providing large-scale computing resources for participating startups.

Meanwhile, AI infrastructure company Firmus Technologies is building a major data center in Batam, Indonesia. Once completed, the facility is expected to scale to 360 megawatts of power capacity and accommodate as many as 170,000 Nvidia GPUs, making it one of Southeast Asia’s largest AI computing hubs.

The expansion highlights Nvidia’s efforts to diversify AI infrastructure geographically as demand for computing power accelerates across Asia-Pacific.

AI industry executives now see GPUs as the “new oil” because computing power has become the critical resource determining which companies can build competitive AI systems.

Demand has grown so rapidly that access to GPUs has become a strategic advantage, with some industry participants reportedly treating compute capacity almost like financial assets through arrangements resembling futures contracts that lock in future access and pricing. Unlike previous technology cycles where capital was often the primary bottleneck, today’s AI startups frequently cite access to computing resources as their biggest challenge.

By allowing startups to exchange future revenue for immediate compute access, Nvidia is addressing one of the sector’s most significant constraints while expanding its own long-term growth opportunities.

Alternative Financing Gains Traction Across AI

Revenue-sharing agreements have become increasingly common as AI companies grapple with enormous infrastructure costs. Training frontier AI models can require investments running into hundreds of millions or even billions of dollars, making traditional financing insufficient for many startups. Instead, developers are increasingly turning to strategic partnerships involving revenue sharing, equity investments, and infrastructure financing.

OpenAI has entered into several strategic agreements involving investments and partnerships with companies including Amazon and AMD as it expands its AI infrastructure. Across the industry, companies are seeking ways to secure long-term computing capacity without placing excessive strain on their balance sheets.

For Nvidia, the model also creates an opportunity to benefit financially from the rapid expansion of AI applications long after its chips have been delivered. Earlier this month, the company disclosed plans to raise debt that sources said could total at least $20 billion. Nvidia said the proceeds would be used for general corporate purposes, including refinancing existing debt, while giving it additional financial flexibility as demand for AI infrastructure continues to surge.

The financing underlines Nvidia’s confidence that AI investment remains in its early stages, with hyperscale cloud providers, governments and enterprises expected to spend hundreds of billions of dollars over the coming years on data centers and advanced computing infrastructure. That spending wave has already transformed Nvidia into one of the world’s most valuable companies and the undisputed leader in AI semiconductors.

The revenue-sharing initiative indicates the company now wants to capture value beyond chip sales by participating directly in the commercial success of AI startups.