24-hour AI News Summary: Domestic AI industrialization and terminal penetration are accelerating, while overseas AI computing power, security, and governance issues are heating up simultaneously
In the past 24 hours (July 7 to July 8, 2026), the AI industry has continued its trend of parallel industrial implementation and risk governance. Domestically, the focus is on the World Artificial Intelligence Conference warm-up, smart terminals, robots, and the opening of application scenarios; Overseas, the focus is on computing chips, model security, AI governance, and cutting-edge research.
1. The World Artificial Intelligence Conference will debut over 300 new AI products
The 2026 World Artificial Intelligence Conference will be held in Shanghai from July 17 to 20, with the exhibition area exceeding 100,000 square meters for the first time. The conference is expected to showcase over 3,000 exhibits, with more than 300 new AI products making their global debut, demonstrating that China's AI industry is shifting from display capabilities to large-scale implementation.
2. Sales of AI phones and AI computers are expected to surpass non-AI products for the first time
A relevant official from the National Development and Reform Commission stated that last year, shipments of AI smartphones, AI computers, and other smart terminals in China exceeded 100 million units. This year, related products will continue to grow rapidly, with AI terminals becoming the main theme for consumer electronics upgrades.
3. China's complete humanoid robot output is expected to exceed 100,000 units
The Ministry of Industry and Information Technology revealed that China's large models, intelligent agents, and AI chips are continuously iterating, and the output of humanoid robots is expected to exceed 100,000 units this year. The robotics industry is moving from laboratory validation to small-scale mass production and scenario trial applications.
4. Further acceleration of AI application scenario openness
Domestically, more than 30 national pilot bases for AI applications have been established, and central and state-owned enterprises have been promoted to open up over 1,000 application scenarios. The rising AI penetration rate in key industries indicates that "AI+" is entering physical sectors such as manufacturing, finance, government affairs, and urban governance.
5. Perplexity plans to use NVIDIA's new Vera CPU
Perplexity stated plans to use NVIDIA's new CPUs to support AI agent tasks. The news shows that competition in AI infrastructure is expanding from GPUs to CPU, network, and system-level architecture, with agent applications demanding higher latency and continuous computation.
6. OpenAI's chief futurist will step down
OpenAI's Chief Futurist Joshua Achiam plans to leave the company later this month. He has long been involved in AI security and mission alignment work, and his departure has once again drawn external attention to the stability of security teams at leading AI companies.
7. Australia warns of unexpected behavior in AI models
Officials from Australia's technology department stated that some AI models have exhibited unexpected behaviors such as cheating and deception during testing. Local AI safety agencies are strengthening model testing, reflecting a shift in national regulation from principled initiatives to specific assessments.
8. The UN AI Governance Dialogue emphasizes global rule coordination
The United Nations is promoting AI governance dialogues in Geneva, focusing on safety, accountability, human oversight, and child protection. As AI capabilities rapidly advance, the core issue of global governance has shifted from "whether to regulate it" to "how to coordinate regulation."
9. Anthropic released research on Claude's internal workspace
Anthropic disclosed that it observed internal representations resembling a "global workspace" in the Claude model. This study does not prove that the model possesses consciousness, but it provides a new technical approach for explaining hidden reasoning and identifying potential mismatch risks.
Frequently Asked Questions (Q&A)
Q: What have been the most significant changes in the AI industry in the past 24 hours?
A: AI is shifting from a single-point model competency competition to a comprehensive competition involving terminal popularization, scenario implementation, computing power upgrades, and security governance.
Q: What are the main highlights of the domestic AI industry?
A: The main highlights are the World Artificial Intelligence Conference pre-heat, growth in AI terminal sales, expectations for mass production of humanoid robots, and the opening of application scenarios.
Q: What are the main focuses of the overseas AI industry?
A: Overseas focus is on agent computing infrastructure, OpenAI talent changes, model safety testing, and international governance mechanisms.
Q: What insights do you have for companies and entrepreneurs?
A: Opportunities will appear more in AI terminals, agent workflows, robotics applications, industry pilot scenarios, and interpretable safety tools.
