In 2025, artificial intelligence (AI) stocks have faced significant challenges despite no indications of the AI boom diminishing. This downturn is attributed to tariff threats and policies from the Trump Administration. Although AI stocks had seen a substantial increase over the past two years, they entered the year with relatively high valuations. The looming threat of an economic recession has raised concerns about the large AI investments projected by major tech companies earlier in the year.
On April 9, the administration reduced the most severe reciprocal tariffs. Simultaneously, several prominent tech CEOs reaffirmed their substantial AI investment plans for the year, highlighting the continued strong demand for AI. However, some comments suggested potential risks for Nvidia, a leading company in the AI sector.
The AI sector appears resilient to recession. Despite market instability, the AI revolution shows no signs of slowing down. Recently, two CEOs from the “Magnificent Seven” companies confirmed this trend. Alphabet hosted its Google Cloud Next 2025 event, where CEO Sundar Pichai announced plans to invest $75 billion in AI data centers this year. Pichai emphasized the expansive potential of AI, and the event featured positive feedback from major customers like Intuit and Verizon regarding Google’s AI models.
The introduction of China’s cost-effective DeepSeek R1 model earlier in the year has also impacted AI stocks. On a positive note, Amazon CEO Andy Jassy clarified in a CNBC interview that reduced costs can drive more innovation, even leading to increased overall spending. In his shareholder letter, Jassy highlighted Amazon’s triple-digit growth rates in AI revenues and expressed optimism about generative AI’s transformative potential.
Concerns about AI costs have emerged as a significant factor, particularly in a potential recession. This situation presents challenges for Nvidia, which is closely associated with AI development, particularly due to high demand for its Blackwell chip. Both Amazon and Google have voiced efforts to reduce AI costs, with Jassy specifically pointing out the need to decrease the expense of AI chips, which are mainly provided by Nvidia.
Nvidia’s chips are notably expensive, and with the company’s substantial gross margins, there is speculation that Nvidia’s revenue and margins may eventually be affected if competitors successfully develop more cost-efficient alternatives. Amazon’s new Trainium2 chip claims to offer better price performance compared to Nvidia’s offerings, and the company is also developing Trainium3. Similarly, Alphabet introduced a powerful in-house chip named Ironwood during their recent event, showcasing its advancements in AI chip technology.
Competition is increasing for Nvidia, which has a lead in AI chip development. Its CUDA software provides a temporary competitive edge. However, companies like Amazon and Google have the capability to manufacture chips at lower costs, potentially impacting Nvidia’s position. If successful in their developments, these companies could lower AI costs and reduce Nvidia’s role in the market.
