24-hour AI news: Google 5 billion pounds in the UK, Tencent Hunyuan 3D 3.0 unveiled

In the past 24 hours, Tencent on the domestic side has launched an upgrade of its agent and 3D generation capabilities in Shenzhen, and the country has released the "Artificial Intelligence Security Governance Framework" 2.0 and announced the results of the first large model crowd test; Overseas, Google announced an investment of £5 billion in the UK and the opening of a data center near London in the next two years, OpenAI brought in former xAI executives, Workday completed a $1.1 billion acquisition, and Figure's new round of financing was valued at $39 billion.

1. Google invests £5 billion in two years to deploy AI and data centers in the UK

1. On September 16, it announced a total investment of £5 billion in the next two years, covering data centers, cloud computing and AI research and development.
2. The Waltham Cross data center in the suburbs of London was opened, with the goal of achieving 95% carbon-free electricity for UK operations by 2026.
3. It is expected to drive an average of about 8,250 jobs per year for British enterprises and strengthen the AI economic ecology.

2. Tencent released ADP 3.0 and Hunyuan 3D 3.0

1. On September 16, the Agent Development Platform (ADP) 3.0 was launched at the Shenzhen Global Digital Ecology Conference.
2. The modeling accuracy of Hunyuan 3D 3.0 is improved by 3 times compared with the previous version, and a one-stop pipeline tool "Hunyuan 3D Studio" has been added.
3. The application scenarios cover games, film and television and industrial design, and promote standardization and cost optimization.

3. Artificial Intelligence Security Governance Framework 2.0 officially released

1. It was released at the main forum of Kunming Cybersecurity Week on September 15 and was formulated by the National Internet Emergency Center.
2. The framework adds a new "risk classification + dynamic governance" mechanism to strengthen the security of the whole process of R&D, deployment, and operation and maintenance.
3. Put forward new standards for the evaluation and compliance requirements of large model red teams, marking the upgrade of supervision.

4. Announcement of the results of the first large model mass test in China

1. The results were disclosed on September 16, mobilizing 559 security researchers covering 15 products.
2. "Unique risks" account for more than 60%, highlighting the problems of confrontation samples, overstepping calls and data leakage.
3. It is recommended to improve model guardrails, red team drills and supply chain security to promote verifiable security.

5. OpenAI appoints former xAI CFO as business finance officer

1. On September 16, it was confirmed that Mike Liberatore would be appointed as Business Finance Officer.
2. Responsibilities cover computing power and capital allocation, and cooperate with the CFO and technical team for cost control.
3. Reflect the personnel upgrade of top AI companies in the competition for capital and computing power.

6. Workday acquires AI company Sana for about $1.1 billion

1. On September 16, it announced the completion of the acquisition, planning to integrate Sana's knowledge management and automation technology.
2. Enhance AI capabilities in HR and financial products to improve customer retention and cross-selling.
3. Reflect the trend of accelerating "mergers and acquisitions" in enterprise-level application fields.

7. Figure's new financing valuation reaches $39 billion

1. The latest financing commitment exceeds $1 billion, with NVIDIA and Intel Capital participating.
2. Focus on "humanoid robot + cloud training + enterprise landing" to promote the commercialization of embodied intelligence.
3. The capital side continues to be optimistic about manufacturing automation and long-term AI value.

8. Meta Connect may release the first smart glasses with display screen (according to media reports)

1. It is expected to be launched on September 17-18, priced at about $800, and supports notifications, translation and navigation.
2. It will be equipped with wristband gesture control and developer toolkits to expand the AI interaction ecosystem.
3. The market focuses on ecological construction and interactive verification, rather than sales.

9. Zhongshan Hospital released the "Metamedical Simulation Laboratory"

1. It was announced on September 16, which belongs to the construction progress of the "National AI Application Pilot Base".
2. The laboratory focuses on virtual and real integration clinical training and six major medical agent experience scenarios.
3. Promote the closed loop of scientific research, pilot testing, and landing, and form a hospital-level demonstration.

10. Mushroom Carlink and Tencent Cloud strategic cooperation to build a car city AI network

1. On September 16, the two sides will jointly deploy Robotaxi, vehicle-road collaboration and simulation applications.
2. Emphasize "low-cost large-scale" high computing power to serve smart transportation scenarios.
3. Help autonomous driving move from demonstration areas to large-scale applications.

Frequently Asked Questions (Q&A)

Q: What is the focus of Google's £5 billion investment?

A: It covers data center construction, Google Cloud and DeepMind research and development, and simultaneously promotes the use of clean energy, with the goal of achieving 95% carbon-free electricity in the UK business by 2026.

Q: What are the core advantages of Tencent Hunyuan 3D 3.0?

A: The accuracy is improved by 3 times, and the integrated process from modeling to release is realized in combination with the "3D Studio" tool, which significantly shortens the production cycle and reduces labor costs.

Q: What are the changes in the AI Security Governance Framework 2.0 compared to 1.0?

A: The new risk classification and dynamic governance mechanism emphasizes red team evaluation and full-process control, which is more adaptable to the complex risks brought by generative AI.

Q: How will Workday's acquisition of Sana impact enterprise customers?

A: Enterprise customers will receive native AI Q&A and automation capabilities in HR and financial processes, reducing the complexity and compliance risks of multi-vendor integration.