In the past 24 hours (March 24 to March 25, 2026), domestic AI topics have focused on the rapid increase in application volume, open source routes and the implementation of agents, while overseas have focused on infrastructure, antitrust review and product line adjustment. On the whole, industry competition is further shifting from model capability competition to landing efficiency, computing power supply and commercial closed-loop.
1. The average daily lemmotype call volume in China exceeded 140 trillion yuan
According to the National Data Bureau, by March 2026, the average daily etymology call volume in the country has exceeded 140 trillion yuan, an increase of more than 40% from the end of 2025. This data shows that the domestic large model call has moved from the trial stage to large-scale production and use, and the pace of commercialization has accelerated significantly.
2. Yang Zhilin said that AI R&D will enter the "AI-led era"
Yang Zhilin, founder of the dark side of the moon, said at the Zhongguancun Forum that more and more research work will be completed by AI in the future. The statement reflects that leading domestic startups are focusing on efficient R&D systems and open source ecological expansion.
3. The popularity of intelligent agents continues to spill over to the offline industrial circle
In Guangzhou-related activities, the installation, demonstration and project roadshow around OpenClaw continued to heat up, and manufacturers such as Baidu, Tencent, and Zhipu were mentioned. It shows that agents are no longer just a topic for developers, but are spreading to a wider range of industry people.
4. AI computing power drives domestic EDA and packaging design upgrades
During SEMICON China, domestic EDA companies proposed system-level design upgrades for the AI era. As large models push up advanced packaging, interconnect, and heat dissipation requirements, AI is pushing chip design competition from single-chip to system collaboration.
5. OpenAI announced the suspension of Sora's independent video platform
OpenAI decided to shut down Sora's standalone application and shift resources to next-generation models and infrastructure. This move shows that although the generated video has attracted attention, the pressure of commercial retention and continuous investment is still there, and leading companies have begun to shrink their non-core product lines.
6. EU antitrust officials meet with executives from OpenAI, Google, Meta and Amazon
The EU is incorporating the AI ecosystem into a more complete competition review framework, focusing on whether chatbots, training data, and cloud infrastructure strengthen platform monopolies. The focus of supervision has shifted from a single product to control the entire industry chain.
7. Microsoft took over the large-scale data center project in Texas
Microsoft will lease a large data center project in Texas that was originally negotiated by Oracle and OpenAI, showing that hyperscale computing resources are still the focus of competition for giants. Whoever can lock in the power, computer room and chip will have a better chance of supporting the next round of model iteration.
8. Arm released new AI chips and raised growth expectations
After Arm launched new AI chips, the market's expectations for its revenue contribution increased significantly. The computing power chain is refocusing on the chip design side, and AI dividends continue to spread to underlying hardware and IP suppliers.
Frequently Asked Questions (Q&A)
Q: What is the biggest industry signal in the last 24 hours?
A: AI competition has shifted from "who has a stronger model" to "who lands faster, has more stable resources, and has more controllable costs".
Q: What are the most prominent features of AI dynamics in China?
A: The rapid growth in the number of calls indicates that enterprises and application scenarios are accelerating their access, and large models are entering a stage of stronger large-scale use.
Q: Why do overseas companies emphasize both regulation and computing power?
A: Because leading models are increasingly relying on huge infrastructure investments, regulators are also beginning to pay attention to whether this concentration of resources will create new platform barriers.
Q: What are the inspirations for developers and entrepreneurs?
A: The space for simply doing general chat products continues to shrink, and there are more opportunities for agents, enterprise workflow transformation, AI infrastructure and vertical scenario applications.
