Nvidia’s once-dominant hold on China’s advanced AI accelerator market has collapsed entirely, with Chief Executive Officer Jensen Huang confirming that the company’s share has plunged from about 95 percent to zero.

Huang made the disclosure during a live interview at the Future of Global Markets 2025 event hosted by Citadel Securities in New York on October 6, describing the loss as an unintended consequence of U.S. export controls.

“At the moment, we are 100% out of China,” Huang said. “We went from 95% market share to 0%.”

The remarks, which were later published in full by Citadel Securities on YouTube, mark the first time Nvidia has publicly quantified the impact of the restrictions. Though Huang did not reference specific products, his comments were clearly directed at Nvidia’s data center GPU business — the backbone of global AI infrastructure and one of the fastest-growing segments of the company’s portfolio.

A Collapse Driven by U.S. Policy

Since Washington began tightening export controls on high-end semiconductors in October 2022, Nvidia has faced repeated disruptions to its product line in China. The company’s China-optimized A800 and H800 chips, designed to comply with earlier restrictions, were later rendered ineligible under expanded rules introduced in 2023. Nvidia’s most recent offering, the H20 GPU, also encountered licensing complications that prevented large-scale sales.

On July 31, China’s Cyberspace Administration summoned Nvidia executives for a closed-door meeting, demanding an explanation over concerns that the H20 chips could contain hidden hardware mechanisms enabling remote shutdown or unauthorized access—so-called “backdoors” that bypass normal security protocols. Such vulnerabilities, if proven, could theoretically allow foreign actors to disrupt AI systems or siphon sensitive data without detection.

In September, the Financial Times reported that the CAC directed companies, including ByteDance, parent of TikTok, and e-commerce giant Alibaba, to terminate their testing and orders of Nvidia chips, including RTX Pro 6000D. The order also covered existing purchases, effectively forcing cancellations.

The cumulative effect has been a near-total halt to shipments of advanced AI accelerators to the Chinese market — a market that, according to Nvidia’s own disclosures, accounted for 20 to 25 percent of its data center revenue before the sanctions. That segment alone generated more than $41 billion in the company’s latest financial year, representing a 56 percent year-on-year increase.

“I can’t imagine any policymaker thinking that’s a good idea — that whatever policy we implemented caused America to lose one of the largest markets in the world to 0%,” Huang said, criticizing the policy outcome rather than the intent.

Strategic Blow to Both Sides

The export controls were designed to prevent China from accessing chips that could accelerate the development of advanced military AI systems. However, the policy’s collateral effects are increasingly visible: Nvidia, the world’s most valuable semiconductor company, has effectively been locked out of one of its largest markets, while Chinese companies have been forced to accelerate their domestic chip development programs.

“In all of our forecasts… we’re assuming 0% for China,” Huang said. “If anything happens in China… it will be a bonus.”

That stark admission points to how deeply the restrictions have reshaped Nvidia’s business strategy. The company has now shifted its focus entirely toward North America, Europe, and the Middle East, where demand for AI infrastructure remains exceptionally strong.

Meanwhile, Chinese hyperscalers and AI research labs — including Baidu, Alibaba, Tencent, and ByteDance — have been investing heavily in indigenous silicon efforts. Domestic chipmakers like Huawei’s HiSilicon and startups backed by Beijing’s state investment arms have gained new momentum, attempting to fill the performance gap left by Nvidia’s absence.

Fragmentation in the AI Supply Chain

The rupture has exposed how fragile the global AI hardware ecosystem has become. Huang himself has repeatedly warned that sweeping export restrictions could “spur the development of competitive substitutes” and create parallel technological spheres — one dominated by the U.S. and its allies, and another centered around China’s domestic innovation drive.

That fragmentation is now unfolding rapidly. China’s AI industry has turned to alternative architectures such as Huawei’s Ascend processors and cloud platforms built on local chips, while simultaneously scaling software optimization to squeeze more performance from limited hardware.

Although U.S. policymakers argue that the curbs are essential for national security, industry analysts caution that the restrictions risk eroding American companies’ long-term leverage in global technology markets. Nvidia’s forced retreat provides an early case study of that effect.

A Market Written Off — For Now

While Huang said he still hopes that Nvidia can eventually resume business in China, the company’s near-term outlook leaves no room for optimism.

The development caps a two-year struggle for Nvidia to navigate the geopolitical crossfire between Washington and Beijing. What began as a temporary workaround strategy through products like the A800 has now become a complete disengagement.

For now, the company’s record profits from surging global demand for AI chips have softened the blow. But the broader implications: the world’s top AI chipmaker being locked out of the world’s second-largest economy, while China, once its biggest growth market, races to build a rival ecosystem of its own, remains profound.

The United Kingdom’s technology sector is cutting back sharply on graduate recruitment, with new figures showing a 46 percent drop in entry-level roles over the past year — and an additional 53 percent decline projected for the coming year.

According to the Institute of Student Employers (ISE), the dramatic contraction signals a major shift in how the industry hires as artificial intelligence begins to take over the very tasks once assigned to new recruits.

The ISE’s annual data reveal that while overall graduate hiring across industries fell by 8 percent — the first annual decline since the pandemic-driven slump of 2020 — the technology and pharmaceutical sectors have been hit hardest. For tech, the change is both structural and symbolic as the sector that built the tools of automation is now being reshaped by them.

Stephen Isherwood, joint chief executive of ISE, said the displacement of entry-level professionals by AI was no longer theoretical.

“It is a tough market for students and young people in general. There is not much churn in the labor market and young people are suffering,” he told the Financial Times.

Employers are finding that AI systems can perform routine coding, data analysis, and digital maintenance — the same basic tasks that used to define the early stages of many tech careers. As a result, companies are choosing to recruit fewer graduates and instead prioritize experienced workers who can integrate AI into higher-value processes.

Despite the decline, tech remains the top field for graduate opportunities by volume, with 46 percent of organizations across the UK economy still seeking IT, digital, and AI skills. Yet the paradox remains that demand for technology skills is growing, even as the sector provides fewer openings for the youngest entrants.

The survey also highlights an ironic twist — while AI is reshaping the job market, it has not yet become central to graduate hiring itself. ISE found that although over half of employers use automated systems to manage some testing processes, direct AI use remains rare. Only 15 percent of employers said they employ AI tools in gamified assessments, though that number is expected to rise as both companies and candidates become more comfortable using AI technologies in recruitment.

Meanwhile, employers are increasingly wary of applicants using AI to gain an unfair advantage. Seventy-nine percent of surveyed companies said they were redesigning or reviewing their recruitment procedures in response to AI developments. Still, only 15 percent said they had never suspected candidates of using AI tools to cheat on assessments.

The broader trend has been visible across the global technology industry this year. Major software firms such as Salesforce and Workday have announced thousands of job cuts while simultaneously expanding their deployment of generative AI to handle internal operations. Microsoft, too, confirmed plans to eliminate about 10,000 positions in its workforce as it integrates new AI tools across its products and services.

The result, according to analysts, is an emerging “AI paradox” in employment — the very technology that fuels innovation in the sector is also squeezing out the entry routes for the next generation of professionals. What used to be considered “learning work” — debugging code, managing data sets, or assisting in development cycles — is now being done faster and cheaper by automation.

Isherwood’s comments suggest that the shift could have lasting implications. With fewer opportunities for graduates to gain early experience, the pipeline for mid-level talent may dry up within the next five years. The short-term efficiency gains companies enjoy today could lead to long-term skills shortages as fewer workers gain hands-on experience in the fundamentals of software engineering and data systems.

The contraction comes at a time when AI literacy is becoming essential in nearly every profession — yet the pathway into the tech sector, where those skills are most relevant, appears to be narrowing.

Experts warn that while firms may be optimizing for productivity, they risk undermining the development of human expertise that drives innovation. Without sufficient junior roles, fewer workers will have the chance to master the skills that senior technical positions require, leaving the sector more dependent on a shrinking pool of experienced professionals.

For graduates, the new reality is that even the most technologically advanced industries are no longer guaranteed to offer a starting point. The same tools that power progress — from automated code assistants to machine learning models — are now competing directly with them.

If ISE’s projections prove accurate, the UK tech sector could be facing not just a hiring slowdown, but a generational gap. And for thousands of students entering the workforce, the question is no longer how to get into tech — but whether there will still be an entry point at all.

Elon Musk’s social platform, X, is rolling out a new Handle Marketplace that allows Premium Plus and Premium Business users to browse and request inactive usernames — a move that effectively turns dormant handles into digital assets for sale.

The X Handle Marketplace is our industry-first solution to redistribute handles that are no longer in use.

Eligible Premium subscribers will be able to search and make requests, with both complimentary and paid options available.

— Handle Marketplace (@XHandles) October 19, 2025

The company is framing the initiative as part of its broader effort to create new revenue streams and attract more paying subscribers.

According to the company’s latest announcement, handles will be divided into two main categories. Priority handles, which include typical full names, multi-word phrases, or alphanumeric combinations, will be available for free to eligible users.
Rare handles, however, will come with a price tag ranging from $2,500 to over seven figures, depending on the handle’s demand and uniqueness.

The launch represents one of X’s most ambitious attempts yet to monetize digital identity on the platform, transforming usernames—once free and available on a first-come, first-served basis—into premium digital real estate.

How the Marketplace Works

X describes the Handle Marketplace not as a one-time perk, but as an ongoing subscription-based service integrated into its Premium tiers. Only Premium Plus and Premium Business subscribers will have access, allowing them to search a catalog of dormant or inactive accounts and request ownership of their usernames.

Once a user secures a new handle through the marketplace, their previous username will be frozen, making it unavailable to others. The company also noted that it may introduce a redirect option—which would forward traffic from a user’s old handle to the new one—as an additional paid add-on in the future.

However, the ability to retain these rare handles will depend on a user’s subscription status. X stated that if a user downgrades or cancels their Premium plan, their account will automatically revert to its original username, and access to the marketplace-acquired handle will be revoked. This condition effectively locks premium usernames behind a continuous paywall.

The decision to monetize usernames follows months of speculation about X’s plans to auction off dormant handles. Musk first hinted in late 2023 that the company would explore reclaiming and reselling inactive usernames as part of efforts to “clean up” legacy accounts.

The marketplace formalizes that concept, aligning with X’s ongoing push to diversify income sources beyond advertising, which has declined sharply since Musk’s takeover. The company aims to attract more high-value subscribers, especially businesses, influencers, and collectors looking to secure distinctive brand identities by introducing an exclusive username economy.

Digital branding experts say this approach could turn social media handles into a form of digital real estate—similar to domain names in the early 2000s—where scarcity and prestige drive value.

A Strategy to Drive Paid Subscriptions

The Handle Marketplace fits neatly into X’s strategy to grow its subscription ecosystem. Since Musk rebranded Twitter to X, the company has leaned heavily on paid plans—Premium, Premium Plus, and Business tiers—as it seeks financial stability amid ongoing advertiser pullback.

Premium Plus users currently pay $16 per month, while Premium Business subscribers are billed $42 per month. These plans offer enhanced features such as reduced ads, higher ranking in replies, and increased revenue-sharing opportunities. The addition of a handle marketplace adds a tangible, status-driven incentive for users to remain subscribed at higher tiers.

Some believe the service could generate substantial non-advertising revenue if priced effectively, particularly among businesses seeking short, recognizable, or brand-aligned handles.

Potential Implications

While the initiative may appeal to businesses and elite users, it also raises questions about fairness and access. Critics argue that placing rare handles behind a paywall could erode X’s long-standing culture of open participation, effectively pricing out regular users from desirable usernames.

There are also concerns about transparency and disputes over ownership, particularly if inactive accounts are reclaimed and resold without clear notification to former holders. In similar past efforts, social media platforms like Instagram and TikTok have faced backlash for reclaiming inactive usernames for brand use or resale.

Additionally, the pricing structure—starting at $2,500 and reportedly reaching seven figures—suggests that X is betting heavily on exclusivity and prestige, mirroring auction-style domain markets rather than traditional social media systems.

A New Phase in X’s Monetization Drive

The Handle Marketplace launch continues X’s broader experimentation with paid digital identity features, following earlier initiatives such as verification badges and ad-revenue sharing for Premium users.

It also reflects Musk’s broader ambition to reposition X as an “everything app,” blending social media with payments, commerce, and professional networking.
X appears to be creating a new digital economy within its ecosystem—one where identity itself becomes a subscription-driven commodity.

For X’s business model, the question now is whether this marketplace becomes a sustainable revenue engine or just another feature that benefits a small fraction of elite users.

Either way, it marks another bold step in Musk’s push to rebuild X’s financial foundation around direct user monetization rather than advertising—a transformation that continues to reshape the platform’s identity, economy, and culture.