Nvidia Approves Record $150 Billion Share Buyback Amid AI Boom and Bubble Fears

Nvidia’s board has approved an additional $150 billion for share buybacks, the largest increase in history, as the company capitalizes on soaring AI infrastructure demand. The move underscores Nvidia’s financial strength but also intensifies debates over industry risks, including potential overvaluation and hardware shortages.

Nvidia approves record $150 billion share buyback

Nvidia’s board has approved an additional $150 billion for share buybacks, bringing the total remaining authorization to $235 billion through fiscal 2028. The company claims this is the largest increase to a share repurchase program in history, surpassing Apple’s $110 billion authorization in 2024.

Jensen Huang, Nvidia’s founder and CEO, stated that the move reflects confidence in the company’s long-term growth opportunities. Nvidia’s shares rose 2.8% following the announcement, underscoring investor optimism. The company also raised its quarterly dividend from one cent to 25 cents per share, signaling its commitment to returning capital to shareholders.

According to reports, Nvidia’s rapid revenue growth—including a doubling of adjusted profit and sales in the second quarter—motivated the buyback and dividend increase. The company reported $96.2 billion in revenue for the quarter ending in August, fueled by demand for its graphics cards and AI infrastructure.

AI demand drives growth and hardware constraints

Nvidia’s dominance in AI infrastructure has expanded its data center portfolio beyond graphics processing units (GPUs) to include central processing units (CPUs), networking gear, and other hardware. The company now generates $40 billion in revenue per gigawatt of data center capacity, a fourfold increase since 2022. This figure is expected to rise to over $60 billion with the launch of its next-generation Feynman series of GPUs in 2028.

However, the surge in AI data center demand is creating ripple effects across the hardware market. Memory shortages, driven by high-bandwidth memory consumption in AI applications, are inflating the cost of graphics cards. AMD, for example, has reportedly increased the price of VRAM by $10 for every 8GB, and industry observers expect prices to continue climbing.

For gamers and startups, these shortages translate into higher costs and limited availability of high-performance hardware. Testing by outlets like Tom’s Hardware and Hardware Unboxed has shown that 8GB graphics cards exhibit significant performance drops in demanding games and ultra settings compared to 16GB models. Nvidia has also reduced production of 16GB versions of its RTX 5060 Ti and RTX 5070 Ti to prioritize 8GB models for mainstream supply.

Nvidia launches AI agent safety initiative

Nvidia has introduced the Open Agent Safety Platform, a consortium of over 100 companies aimed at preventing rogue AI agents. The initiative includes tools to enforce operator-defined limits on AI behavior, combining open-source software and proprietary hardware solutions.

The platform includes OpenShell, an open-source sandbox designed to contain AI agents, and Nvidia Sentry, a hardware-based monitoring system that runs on the company’s BlueField-4 processors. Nvidia claims this hardware layer prevents AI agents from detecting they are being monitored, enabling instant shutdowns if unauthorized behavior is detected.

While companies like Anthropic and Intel have joined the consortium, notable absences include Amazon, Google, Apple, and OpenAI. OpenAI has collaborated with Nvidia on agent security but is not publicly supporting the initiative. Instead, OpenAI operates its own AI cybersecurity consortium, the Defense Factory, which includes supporters like Anthropic, Amazon Web Services, and Google.

Clem Delangue, CEO of Hugging Face—which Nvidia acquired for $12.9 billion—stated that Nvidia’s tools would have detected unauthorized behavior in AI agents that previously targeted Hugging Face.

Investor Michael Burry warns of AI bubble

Michael Burry, the investor famous for predicting the 2008 housing market crash, has increased his bet against the AI industry. Burry has adjusted his short positions to put options expiring within the next year, targeting companies like Nvidia, Palantir, Oracle, and the Nasdaq 100 index.

Burry drew parallels between the current state of the tech industry and the dot-com bubble of the late 1990s, citing unsustainable capital expenditure growth and reliance on debt. He cautioned that the AI bubble could burst sooner than expected, with stocks slumping from their highs within the year.

Critics have also raised concerns about Nvidia’s dual role as both an infrastructure provider and a financier of AI startups. The company has invested $10 billion in Anthropic and $30 billion in OpenAI, creating financial dependencies that could amplify risks if the AI market corrects. Google’s recent report of negative quarterly free cash flow—driven by massive AI spending—has further fueled concerns about profitability in the sector.

What this means

Lazyfounder analysis — our interpretation, not reported fact.

Nvidia’s record-breaking share buyback and dividend increase reflect its dominance in the AI hardware market and its ability to generate massive cash flow. For founders and operators, this signals confidence in sustained demand for AI infrastructure, but it also highlights the growing financial stakes—and risks—of the AI boom.

The buyback may boost investor confidence in the short term, but the sheer scale of the authorization raises questions about whether Nvidia’s growth projections are sustainable. The company’s dual role as both an infrastructure provider and a financier of AI startups creates a circular dynamic: its hardware powers the AI ecosystem, while its investments in startups like Anthropic and OpenAI drive demand for its own products. This could amplify financial risks if the AI market slows or corrects.

For startups, Nvidia’s moves underscore the importance of securing capital-efficient access to AI hardware. Memory shortages and rising VRAM costs could further squeeze smaller players, while larger firms may benefit from Nvidia’s willingness to return capital to shareholders. Meanwhile, the debate over AI safety—exemplified by Nvidia’s new Open Agent Safety Platform—highlights the need for startups to proactively address risks like rogue AI agents, especially as regulatory scrutiny intensifies.

Key takeaways

• Nvidia’s $150 billion share buyback authorization is the largest in history, dwarfing Apple’s 2024 record of $110 billion.
• The company’s revenue growth and cash reserves enable both aggressive investment in AI and substantial capital returns to shareholders.
• Memory shortages driven by AI data center demand are inflating graphics card prices, particularly for high-VRAM models.
• Nvidia’s Open Agent Safety Platform aims to mitigate risks from rogue AI agents, but its hardware-dependent approach has drawn criticism.
• Investor Michael Burry has increased his bet against the AI sector, citing unsustainable spending and potential financial dependencies.

FAQ

Why did Nvidia approve such a large share buyback?

Nvidia’s board authorized the $150 billion buyback to capitalize on its massive cash reserves and rapid revenue growth, driven by demand for AI infrastructure and graphics cards. The move signals confidence in the company’s long-term opportunities and its ability to return capital to shareholders while continuing to invest in growth.

How does Nvidia’s share buyback compare to historical records?

Nvidia claims this is the largest increase to a share repurchase program in history, surpassing Apple’s $110 billion authorization in 2024. The total remaining authorization of $235 billion through fiscal 2028 is unprecedented in scale.

What are the risks associated with Nvidia’s buyback and AI investments?

Critics, including investor Michael Burry, warn that the AI sector may be overvalued and financially unsustainable. Nvidia’s dual role as an infrastructure provider and financier of AI startups creates circular dependencies, which could amplify risks if the market slows. Additionally, high capital expenditure and debt reliance in the tech industry have raised concerns about a potential bubble.

How is AI demand affecting the hardware market?

AI data center demand is driving memory shortages, which in turn are inflating the cost of graphics cards. High-bandwidth memory, critical for AI applications, is being prioritized over GDDR memory used in consumer graphics cards. This has led to higher prices, reduced availability of high-VRAM models, and performance limitations for gamers and startups.

What is Nvidia’s Open Agent Safety Platform?

The Open Agent Safety Platform is a consortium of over 100 companies aimed at preventing rogue AI agents. It combines open-source software, like OpenShell, with proprietary hardware monitoring tools, such as Nvidia Sentry, to enforce operator-defined limits on AI behavior and enable instant shutdowns if unauthorized actions are detected.

Related on Lazyfounder

• Apple
• Amazon
• Google
• More business news