24-Hour AI News: UK Locks Down Search Regulation, Trillion-Point Parameters Open Source, and Multimodality on-Device Launches

In the past 24 hours, Ant Group released a trillion-parameter open-source model, Vivo launched a large, multimodal model on-device and upgraded its system strategy. Meanwhile, AI chip inspections at ports have tightened, and applications in government and public services continue to expand. Internationally, UK regulators designated Google a "strategic market position," global financial regulators strengthened AI risk monitoring, competition between OpenAI and the EU intensified, and former British Prime Minister Rishi Sunak became an advisor to Microsoft and Anthropic. Maintaining a balanced coverage of domestic and international news.

1. UK regulators designate Google as having a “strategic market position” for search and search advertising

1. On October 10, the UK Competition and Markets Authority confirmed the implementation of the "strategic market position" designation for Google, covering search and search advertising services.
2. Regulators plan to hold consultations on measures such as "selection screens", non-discriminatory rankings, content use and AI summary annotation.
3. This designation is the first of its kind under the new digital regulatory framework and may impact the product form and competitive landscape of AI search and content ecology in the UK.

2. Ant Group releases and open-sources Ling-1T, a trillion-parameter universal language model

1. Reports from October 9th to 10th showed that the parameter scale of Ling-1T is about 1 trillion, it is positioned as a general-purpose LLM and announced to be open source.
2. Official and media reports indicate that the system leads in benchmarks such as code generation, competitive mathematics, and complex reasoning.
3. Simultaneously form a "non-thinking/thinking/multimodal" spectrum to promote the expansion of China's open source big model ecosystem.

3. Vivo releases Blue Heart 3B end-to-end multimodal large model and system strategy upgrade

1. On October 10, a 3B-parameter edge-side multimodal model was released, supporting approximately 128K contexts. It was claimed that its comprehensive performance within the 10B scale topped multiple rankings.
2. The press conference simultaneously launched system and framework upgrades to promote the deep integration of "AI + operating system".
3. Long-terminal context and multimodal reasoning on the end provide support for low-latency and privacy-friendly generative interactions on mobile phones.

4. China strengthens customs inspections of AI chips at major ports

1. According to news on October 10, customs has strengthened inspections of semiconductor goods, especially AI-related chips, at major ports.
2. The inspection scope is expanded from specific models to a wider range of high-end chips and potential smuggling risk categories.
3. Against the backdrop of global supply and regulatory changes, this move may have a temporary impact on the delivery pace and cost of high-end GPUs in non-US markets.

5. Global financial regulators intend to strengthen risk monitoring of AI use

1. On October 10, multiple media outlets quoted regulatory trends as saying that AI applications in banks and capital markets will be included in more frequent and penetrating monitoring.
2. Focus on model homogeneity, black box decision-making, data leakage and systemic spillover risks.
3. Put forward more detailed requirements for information disclosure, model governance and emergency plans.

6. OpenAI raises competition concerns with EU antitrust authorities

1. Reports from October 9th to 10th indicated that OpenAI communicated with EU officials about the competitive impact of large platforms in the field of AI.
2. The topic involves the distribution advantages and market power assessment of vertically integrated platforms.
3. These discussions are intertwined with the EU’s progress in governing digital competition and generative AI with large technology platforms.

7. Rishi Sunak appointed as advisor to Microsoft and Anthropic

1. It was disclosed on October 9-10 that the former British Prime Minister was approved to serve as a consultant to Microsoft and Anthropic, subject to the constraints of the ACoBA.
2. Their responsibilities focus on global strategies and macro issues, and their salary commitments include donations to public welfare projects.
3. The incident reflects the boundary management between governance and compliance among government, business and cutting-edge AI companies.

8. Beijing held the "AI+Education Innovation and Practice Conference", and "Jingxiaobao" answered about 210,000 questions in total.

1. The conference on October 10 focused on AI-enabled education governance and talent cultivation, with approximately 500 representatives attending.
2. Beijing’s intelligent human resources and social security policy assistant “Jingxiaobao” has been online since July, with a total of about 93,000 visits and nearly 210,000 answers to questions.
3. The large-scale application of government AI assistants in people's livelihood service scenarios continues to advance.

Frequently Asked Questions (Q&A)

Q: What direct changes will the UK’s “strategic market position” designation for Google bring?

A: Regulators will initiate consultations and may introduce measures such as a "selection screen", non-discriminatory rankings, and AI summary annotations; in the short term, there will be rule discussions and pilots, and in the medium term, competition in AI search portals may be reshaped.

Q: What is the practical significance of the trillion-parameter open source model?

A: Raise the upper limit on complex reasoning, coding and mathematical tasks; open source strategy reduces the trial and error costs of scientific research and industry, driving secondary innovation and ecological prosperity.

Q: Why is the large-scale end-to-end model important?

A: The end side has the advantages of low latency and privacy. Combined with long context and multimodal reasoning, it can achieve a more consistent and personalized local intelligent experience on mobile phones and wearable devices.

Q: What are the main regulatory concerns about AI in the financial system?

A: The "herd effect" caused by model homogeneity, the stability and explainability issues brought about by black box decision-making, data compliance and leakage risks, and contingency plans and information disclosure in the event of major errors.