The last two weeks have been intense for the tech sector. From Microsoft (MSFT) to Nvidia (NVDA), everyone appears to be racing to secure market leadership. For big techs and AI firms, it is a good time for new investments.

In the last 15 days alone, Nvidia has closed a historical deal of US$5 billion with Intel (INTC) for the development of AI chips, announced a US$100 billion investment in ChatGPT’s OpenAI, while Microsoft has increased its activity in the AI sector with a strategic partnership with Anthropic, a direct rival of OpenAI’s.

Last year, private investment in AI reached around US$252 billion worldwide, with the United States leading with 109,1 billion in 2024, significantly surpassing countries like China (US$9.3 billion) and the UK (US$4.5 billion), according to the consultancy company Grand View Research.

According to Stanford HAI, the university’s branch, which promotes research, education, politics, and AI practices for the betterment of the human condition, the economic impact of AI could reach US$20 trillion by 2030. This projection suggests that the global AI market, valued at US$391 billion in 2025, is expected to grow to US$1.81 trillion by 2030, with a compound annual growth rate of 35.9%. The growth, according to the institution, will be driven by the increasing adoption of the technology by sectors such as healthcare, manufacturing, finance, and public services.

And investments shouldn’t stop here.

Nvidia and OpenAI: landmark US$100 billion deal

This week, Nvidia announced a US$100 billion investment in OpenAI, expanding the semiconductor giant's role.

The deal, according to the companies, aims to strengthen Nvidia’s position as the leading supplier of graphics cards, essential to the infrastructure behind artificial intelligence. The transaction happens in a strategic moment for both companies, which are increasingly dominating the AI sector.

The analyst Matt Britzman, from Hargreaves Lansdown, highlighted that with this deal, Nvidia guarantees a crucial role in the development of AI data centers.

Britzman estimates that, with each gigawatt of AI data center capacity worth around $50 billion in revenue, this project could generate up to $500 billion in the long term, according to Reuters. The outlook is that the investment in OpenAI will be highly profitable for Nvidia and will consolidate its chips as the backbone of AI infrastructure.

But not all of Wall Street has had a good outlook on the deal. Stacy Regson, from Bernstein Research, has alerted Bloomberg that the transaction may reflect a “circular economy,” where investment among companies becomes excessively dependent on one another, with no effective financial return.

According to the analyst, the sheer size of the OpenAI investment “seems to surpass all others”. In a note to the market, she affirms that the 100 billion dollars “will likely intensify worries more than we have seen before and (maybe rightfully so) raise questions about the justifications behind this action”.

To Rasgon, Nvidia’s investment might be seen as “self-referencing” financing, since both companies are already intimately connected in the AI ecosystem.

Daniel O’Regan, from Mizuho Securities, has also approached the deal critically, suggesting that, despite the market sentiment being generally positive, there is concern that the deal might be viewed as supplier financing, rather than a robust strategic investment. However, the transaction has been widely positive for Nvidia’s shares, which registered a 2.61% increase in the pre-market after the announcement.

Nvidia and Intel: a US$5 billion partnership

Besides the heavy investment in OpenAI, Nvidia has made another critical deal with Intel, worth US$5 billion. The transaction involves the purchase of Intel’s actions, with a discount of 6.5% on last Wednesday’s closing price. The deal aims to combine Nvidia’s graphics technology with Intel’s processors, thus creating new products for data centers and PCs.

The collaboration was seen positively by analysts such as Stifel’s Ruben Roy.

Roy highlighted to Yahoo Finance that the main benefit for Intel will be the strengthening of its position within the AI infrastructure, particularly in the manufacturing of high-density chips for data centers. According to Roy, the move puts Intel, alas, in a good strategic position in the growing AI market.

But Citi’s Christopher Danely has taken a more skeptical stance. The analyst has told CNBC that, although the partnership is meaningful, it won’t necessarily make Intel’s CPUs more competitive in the market, since the integration of Nvidia’s graphics cards with Intel’s processors won’t, by itself, give Intel’s product an edge over the competition.

The transaction occurs at a critical moment for Intel, which is facing difficulties in recovering its leading position in the semiconductor market, particularly in the AI chip market, where Nvidia has demonstrated superior performance.

For analysis by Hendi Susanto of Gabelli Funds, Intel’s success will depend on its ability to secure solid commitments from clients using its 14A technology. Susanto emphasized to MarketWatch that, without these compromises, Intel might face difficulties in maintaining its market position. The Nvidia partnership is seen as an opportunity for Intel to position itself as a major player in AI infrastructure – a lifebuoy amidst chaos.

Microsoft and Anthropic: diversification in Microsoft 365 Copilot

Microsoft has also made a significant move by announcing a partnership with Anthropic, a direct competitor to OpenAI.

From this week onwards, Microsoft 365 Copilot users may alternate between AI models by OpenAI and Anthropic. Offering more options for tasks such as digital research assistance and the creation of custom AI tools. The addition of Anthropic’s models represents a meaningful diversification to Microsoft’s AI portfolio, which has primarily used OpenAI’s technologies so far.

Charles Lamanna, president of Microsoft’s Business and Copilot division, emphasized that the diversification is a strategic move, which offers users more options for AI models, amplifying Copilot’s capabilities and flexibility.

Microsoft had previously invested heavily in OpenAI, with a partnership that granted the company exclusive access to OpenAI's integration rights in its products. Now, Anthropic’s entrance into Microsoft’s AI portfolio might generate tensions in their ties with OpenAI, especially at a moment when OpenAI is seeking more independence from Microsoft.

Is a bubble coming?

The recent billion-dollar movements in the technology sector, especially in the field of AI, have sparked a growing debate about the possibility of a market bubble.

With massive investments from companies like Nvidia, Microsoft, and Intel, the rapid growth in the AI area raises questions about the sustainability of the high market values. Deutsche Bank issued a significant warning, highlighting that AI growth needs to be “parabolic” to justify the high investments, which is considered highly unlikely. Although AI investments are substantial, returns remain low.

For every $1 spent on AI hardware and data centers, only between 30 and 50 cents in annual returns are being generated. According to analysts, the capital expenditures of major tech companies account for 6-8% of the United States' Gross Domestic Product (GDP), surpassing spending on sectors such as automobiles and residential construction. For the bank, this raises the question: how long can this unsustainable growth continue?

Amid these discussions, Mark Zuckerberg, CEO of Meta, and Sam Altman, of OpenAI, also acknowledged the possibility of an AI bubble.

“There are compelling arguments for why AI could be an exception [to market bubbles],” Zuckerberg said on the Access podcast. “And if the models keep growing in capacity year after year and demand continues to grow, then maybe there won't be a collapse.”

However, according to the billionaire, capital spending bubbles, such as the construction of AI infrastructure, primarily seen in the form of data centers, tend to end in similar ways. "I definitely think there’s a possibility, at least empirically, based on past large infrastructure builds and how they led to bubbles, that something like this could happen," Zuckerberg said.