AI stocks worth investing in by 2025

Analysis of the current trend of the artificial intelligence (AI) sector of U.S. stocks and the downside buying strategy

 1. The overall trend and technical support of the AI sector

1.  Pullback background and support analysis
   
   At present, the AI sector of the U.S. stock market is in a stage of adjustment. According to technical analysis, the AI sector has 3 consecutive negative lines on the weekly line, and the trading volume has shrunk, but the 20-week moving average has formed an effective support, and the monthly line shows that it has risen from 546 points to 1714 points and then stepped back, only 40 points away from the doubling target, indicating that there is still a possibility of new highs in the future. The Science and Technology Innovation Artificial Intelligence Index rebounded after a correction of about 7% in March 2025, and the ETF turnover exceeded 100 million yuan, indicating a strong willingness to deploy funds on dips.

2.  Core drivers

   • Policy dividends: The 2025 government work report for the first time listed “artificial intelligence full-chain empowerment” as a key task, covering three major directions: computing infrastructure, model breakthroughs, and application integration.
   • Technology iteration: Falling computing costs, on-device AI breakthroughs (e.g., smartphones, wearable devices), and higher-than-expected capital expenditures are driving long-term growth in the industry.
   • Performance differentiation: The computing power side (such as Nvidia and TSMC) remains strong, and the application side is adjusted due to the lack of commercialization closed-loop, but it is expected to usher in an explosion in 2025.

 2. Current AI segments and individual stock recommendations that are worth paying attention to

 (1) Core computing power and infrastructure layer

1.  Nvidia (NVDA)

   • Technical barriers: GPU market share exceeded 80%, revenue increased by 122.4% year-on-year in 2024Q2, and data center business was the core growth engine.
   • Technicals: The RSI and OBV indicators show a weak bullish divergence, the MACD shows a short-term momentum repair, and a break through the $148 resistance level is expected to resume the rally.
   • Institutional Rating: Jefferies ranks it as the top of the “Big Seven Tech Giants” with a price target of $405 and is bullish on the sustainability of AI computing power demand.

2.  TSMC (TSM)

   • Industrial chain status: the world’s largest semiconductor foundry, the 3nm process exclusively accounts for Apple and Nvidia orders, and the capital expenditure will exceed 30 billion US dollars in 2025.
   • Financial health: In Q4 2024, the gross profit margin will be 58%, the net profit margin will be 42%, and the cash flow will be abundant to support R&D investment.

3.  Microsoft (MSFT)

   • Ecological integration: Investing $13 billion in OpenAI, the penetration rate of Azure AI cloud services has increased, revenue in 2024Q3 has increased by 14% year-on-year, and net profit has reached $48.78 billion.
   • Risk warning: Jefferies believes that its growth momentum is weaker than that of other giants, and it needs to pay attention to the progress of AI agency commercialization.

 (2) Application layer and tool chain

1. Adobe（ADBE）

   • Innovative products: Firefly’s generative AI tools improve design efficiency, with a 15% year-on-year increase in subscribers and a PE of 32x in 2025, lower than the industry average.
   • Institutional Opinion: Huaxing raised its price target to $490, but is bullish on its monopoly position in the creative software space in the long term.

2. Salesforce（CRM）

   • Enterprise AI Transformation: Einstein AI platform has a customer retention rate of over 90%, and Wedbush raised its price target to $375, bullish on the convergence of CRM and AI.
   • Risk: DA Davidson warns that valuations are overvalued and quarterly profit guidance needs to be watched.

 (3) Edge AI and hardware innovation

• Apple (AAPL): The M4 chip integrates a neural network engine, and the iPhone 17 may be equipped with an AI assistant on the device, and the shipment forecast for 2025 has been raised to 250 million units.
• Tesla’s (TSLA) :D ojo supercomputer to accelerate autonomous driving training, and the commercialization of the Optimus robot may become a catalyst, but it is necessary to be wary of performance fluctuations.

 3. Risks and operational suggestions

1.  Short-term risks

   • Valuation pressure: The PE of leading companies is generally higher than the historical median (e.g., Nvidia PE 56.3x), and technical corrections need to be guarded against.
   • Policy uncertainty: The U.S. election may affect the regulation of the technology industry, and Trump’s policy proposition may impact the supply chain.

2.  Position opening strategy

   • Buy in batches: Prioritize leading stocks that have pulled back to key support levels (e.g., Nvidia $140, Microsoft $350) to avoid chasing higher.
   • Decentralized allocation: computing power (40%) + application (30%) + edge AI (30%), balancing risk and return.

 Fourth, the long-term outlook of the industry

The global AI market is expected to grow from $214.6 billion in 2024 to $1.3 trillion in 2030, with a compound growth rate of 35.7%. Generative AI, hybrid architecture (cloud + edge), and vertical industry penetration (healthcare, finance) are the core growth points. Leading companies will continue to lead the way with technical barriers and capital advantages, but application-layer companies need to achieve profitability verification in 2025 to break through the valuation bottleneck.

Conclusion: The current pullback in the AI sector provides long-term investors with layout opportunities, and it is recommended to focus on NVIDIA, Microsoft, TSMC, and Adobe, and closely track the inflection point of application layer performance. Investors with a higher risk tolerance can appropriately allocate to elastic targets such as C3.ai (AI), but need to be wary of the uncertainty of their earnings.