Policy desk

Research institutions need to increase AI R&D investment, particularly in foundation models and intelligent scientist systems

Companies must strengthen core technology R&D in multi-sensor fusion, algorithms, solid-state batteries, automotive chips, and operating systems

Dutch government will gradually reduce reliance on Microsoft products, shifting to NixOS-based autonomous solutions

U.S. AI companies must adjust model release processes to ensure U.S. agencies review first

AI companies must adjust model release processes to ensure U.S. government gets priority testing access

AI elevated to strategic priority level, indicating future policy support and funding

Provides funding support for AI evaluation research and promotes academic collaboration

Mobile on-device AI service providers must complete registration to operate in China

Provides traceability for AI-generated content, helping identify content sources

Manufacturing enterprises are required to accelerate the application of AI in R&D design, production manufacturing, and management operations to enhance digitalization and intelligentization, aligning with national industrial policy directions.

Obligations: Enterprises must invest in building an organizational ontology to provide context for AI agents. • Opportunities: A well-defined ontology enables more effective AI-driven automation and decision-making. • Deadlines: While not explicitly stated, the guide implies a need for immediate action to catch up with the pace of AI-driven network attacks.

Companies must strengthen AI-manufacturing integration, developing smart factories and intelligent manufacturing solutions

Obligations: AI companies may face legal scrutiny and potential penalties if found to have coordinated to limit AI development speed, violating antitrust laws. • Opportunities: The case highlights the importance of transparent communication and compliance with antitrust regulations in the AI sector. • Deadlines: The court case is ongoing, with the possibility of a collective class-action suit being certified, which could impact the companies' operations and public perception.

Obligations: AI companies operating in California may be subject to new regulatory requirements, including potential mandatory security audits by independent third parties. • Opportunities: Companies may gain clarity in the regulatory framework for AI safety, potentially reducing uncertainty. • Deadlines: Companies should prepare for the upcoming two-month timeline for the expert panel to submit its recommendations and the subsequent legislative process.

Obligations: Companies must comply with new evaluation standards for token factories, focusing on core technical metrics like model adaptation capacity and energy efficiency. • Opportunities: Firms with leading token factories can receive government support. Companies involved in model technology iteration, chip adaptation, and overseas service development can tap into new market opportunities. • Deadlines: The plan is set for implementation from 2026 to 2028, with specific milestones for building factories, launching platforms, and achieving technology breakthroughs.

Companies must submit details of agreements with Apple to the court, potentially exposing commercial secrets

Obligations As a model provider, OpenAI has a responsibility to publicly disclose findings of model misalignment and establish clear standards and timelines for tracking, investigating, and disclosing such issues.

Obligations: Content creators must use the self-declaration mechanism to mark content sources. Companies deploying AI-generated content must ensure traceability. Users must adhere to rational and friendly expression guidelines, and avoid online harassment.

Obligations: Companies in the electronics manufacturing sector must align their R&D and production strategies with the plan's goals, particularly in developing advanced chips and promoting domestic operating systems.

Obligations: AI developers must ensure models adhere to the principles, cannot be designed to mimic consciousness, and must not be given legal personhood. Models must be within human control and fail tasks that violate the principles.

Obligations: Universities must adopt and implement the eight principles to redefine educational goals and assessment methods. Teachers are required to establish clear AI policies for each course and be transparent about their own AI usage.

Tech companies must take higher responsibility, needing to strengthen content moderation and risk control to avoid child harm from negligence• Violations will face high fines, increasing compliance costs and risks• Companies must immediately assess existing systems to ensure compliance with the new regulations after the law takes effect

Obligations: Software companies are required to participate in the initiative, develop intelligent software products, and cooperate with terminal manufacturers for joint innovation. • Opportunities: Companies can benefit from government funding and support to develop new intelligent software products and participate in the creation of benchmark applications. • Deadlines: By 2028, the initiative aims to cover 20,000 large-scale software enterprises and create 100 benchmark applications.

Obligations: AI developers and deployers in California must comply with new state-level assessment and audit requirements. • Opportunities: Companies can gain a competitive advantage by demonstrating compliance with these pioneering AI safety standards. • Deadlines: Companies must prepare for potential implementation of the new assessment and audit framework, which is now law.

Requires large AI companies to increase transparency

From 1 July 2027, L3/L4 automated driving systems must meet the mandatory safety requirements; non-compliant models cannot be sold.