Global AI Newsletter·Issue 19

On April 2, the Ministry of Industry and Information Technology issued the Measures for Ethical Review and Services of Artificial Intelligence Science and Technology (Trial) (hereinafter referred to as the "Measures"). The Measures aim to promote the sound development of artificial intelligence technology, ensure that AI scientific and technological activities comply with ethical requirements, and safeguard social and public interests.

The Measures stipulate the following contents: First, the subject of ethical review. The Measures require institutions of higher learning, scientific research institutions, enterprises and other entities to establish artificial intelligence science and technology ethical review bodies, which are responsible for the review and supervision of relevant activities. Second, ethical review principles. The Measures clarify that artificial intelligence science and technology activities shall follow six ethical principles: promoting human well-being, respecting the right to life, fairness and impartiality, reasonable risk control, openness and transparency, and privacy protection. Third, review contents. The Measures stipulate that ethical review shall focus on artificial intelligence training data, rationality of algorithm design, system risks, privacy and security, etc., to prevent algorithm discrimination and other ethical risks. Fourth, the development of the service system. The Measures propose to establish and improve the standard system for artificial intelligence science and technology ethics, promote the development of services such as risk monitoring and early warning and technical support, and enhance the ability to prevent and control ethical risks.

Link: https://www.miit.gov.cn/jgsj/kjs/wjfb/art/2026/art_2995f16b28504ddcbb604e918eb15759.html

2. Cyberspace Administration of China Issued the Measures for the Administration of Digital Virtual Human Information Services (Draft for Comments)

On April 3, the Cyberspace Administration of China issued the Measures for the Administration of Digital Virtual Human Information Services (Draft for Comments) (hereinafter referred to as the "Administration Measures"). The Administration Measures consists of 5 chapters and 27 articles, aiming to promote the sound development and standardized application of digital virtual human information services.

The Administration Measures stipulate the following contents: First, definition. A digital virtual human refers to a virtual digital image that exists in the non-physical world, uses technologies such as graphics, digital image processing or artificial intelligence, is driven by real people or computing, simulates human appearance, and has characteristics such as voice, behavior, interaction ability or personality. Second, protection of natural persons' rights and interests. The Administration Measures stipulate that service providers shall obtain separate consent from natural persons and inform them of the purpose and necessity of processing in a prominent manner when using sensitive personal information of natural persons for virtual human modeling, image generation and other activities, which complies with the "notice-consent" provisions stipulated in the Personal Information Protection Law. Third, service specifications. The Administration Measures clarify that service providers and individuals shall not engage in illegal activities such as generating content that has adverse effects on the online ecology when providing and using digital virtual human services, shall affix a prominent mark containing the words "digital human" throughout the process, and shall fulfill data security protection responsibilities and carry out legal and compliant data processing activities. Fourth, supervision and inspection and legal liability.

Link: https://www.cac.gov.cn/2026-04/03/c_1776952992709096.htm

On April 3, the Cyberspace Administration of China issued the Provisions on Simplified Measures for Personal Information Protection of Small-scale Personal Information Processors (Draft for Comments) to solicit public comments, with the deadline for comments being May 3, 2026.

The draft for comments clarifies that a small-scale personal information processor refers to an entity that processes personal information of less than 100,000 individuals, aiming to support the innovative development of micro, small and medium-sized enterprises and reduce their compliance costs for personal information protection. The main contents include: First, simplifying information processing rules, allowing disclosure through simple methods such as posting announcements and service agreements, and supporting parks, commercial properties and others to uniformly formulate rules for similar businesses. Second, simplifying the notice and consent mechanism. Eligible processors may fulfill the notice obligation only through public rules, and those that process information only through online platforms and comply with platform rules may be exempted from formulating separate rules and conducting repeated audits. Third, relaxing requirements for cross-border data transfer, listing multiple scenarios exempted from declaration such as cross-border shopping, human resource management, and emergency situations. Fourth, simplifying compliance audits, which can be conducted at least once every five years by means of a self-inspection form. Fifth, clarifying a fault-tolerance mechanism, where no penalty or a mitigated or lighter penalty shall be imposed for minor violations that are corrected in a timely manner, first-time violations with minor harm, etc. In addition, all regions and departments are encouraged to help small-scale processors improve their protection capabilities through training, technical tools and other means.

Link: https://www.cac.gov.cn/2026-04/03/c_1776865011603509.htm

4. National Medical Products Administration Issued the Implementation Opinions on "Artificial Intelligence + Drug Regulation"

On April 2, the National Medical Products Administration issued the Implementation Opinions on "Artificial Intelligence + Drug Regulation" (hereinafter referred to as the "Opinions") to promote the innovative application of artificial intelligence in the whole life cycle supervision of drugs, which includes four parts: general requirements, key supervision scenarios empowered by digital intelligence, basic support for "artificial intelligence + drug regulation", and organization and implementation. In the part of key supervision scenarios empowered by digital intelligence, the Opinions propose to build a human-machine collaborative intelligent review and approval system, improve the intelligent supervision capability of the whole chain, promote the digital intelligence upgrade of the risk supervision system, promote the intelligent and standardized inspection and law enforcement, improve the efficiency of collaborative supervision, enhance the intelligent level of government services, and promote the coordinated development of supervision and industrial digital intelligence. In the part of consolidating the basic support for "artificial intelligence + drug regulation", the Opinions point out: First, promoting the construction of high-quality data sets for drug regulation; second, strengthening the support system for artificial intelligence applications; third, enhancing the construction of computing power infrastructure; fourth, building a solid security protection system; fifth, improving the construction and operation management mechanism.

Link: https://www.nmpa.gov.cn/xxgk/fgwj/gzwj/gzwjzh/20260402091552114.html

5. Ministry of Industry and Information Technology and Eight Other Departments Issued the Action Plan for Promoting the Innovative Development of the Internet of Things Industry (2026—2028)

On March 31, the Ministry of Industry and Information Technology and eight other departments jointly issued the Action Plan for Promoting the Innovative Development of the Internet of Things Industry (2026—2028) (MIIT Lian Ke [2026] No. 65). The plan deploys five key tasks with a total of 16 specific works: First, promoting the innovation and upgrading of equipment, making breakthroughs in high-end sensor technology, enhancing the connection capability of network equipment, and promoting the in-depth integration of AI, 5G, edge computing and terminals. Second, improving the efficiency of platform services. Third, cultivating application scenarios, covering production fields such as intelligent manufacturing, smart agriculture, smart grid and smart transportation, consumer fields such as smart home, Internet of Vehicles, telemedicine and digital education, and supporting social governance fields such as digital twin cities and intelligent management and control. Fourth, consolidating the network foundation. Fifth, fostering the industrial ecosystem, including improving the standard system, highlighting the dominant position of enterprises in innovation, deepening the integration of the Internet of Things and AI, releasing data value and strengthening security supervision. The safeguard measures cover overall coordination, the layout of scientific and technological projects, public services, resource support and the construction of talent teams.

Link: https://www.miit.gov.cn/zwgk/zcwj/wjfb/tz/art/2026/art_9d8c75f8355a4179abad5a6296273dd2.html

6. National Information Security Standardization Technical Committee Issued the Practical Guide to Cybersecurity Standards – Maturity Model of Data Security Capability for Industrial Enterprises

On March 31, the National Information Security Standardization Technical Committee issued the Practical Guide to Cybersecurity Standards – Maturity Model of Data Security Capability for Industrial Enterprises (hereinafter referred to as the "Practical Guide"). The Practical Guide puts forward a maturity model of data security capability for industrial enterprises, stipulates the maturity level requirements for the whole life cycle data security and general security of industrial enterprises, and is applicable to guiding industrial enterprises to carry out data security capability construction and evaluating the maturity level of data security capability of industrial enterprises.

Link: https://www.tc260.org.cn/portal/article/2/0604dc4de4e842cd92636c30c43aaa55

On March 31, the National Information Security Standardization Technical Committee solicited public comments on the Practical Guide to Cybersecurity Standards – Security Guidelines for Deployment and Use of OpenClaw-like Agents (Draft for Comments) (hereinafter referred to as the "Guidelines"). The Guidelines provide detailed guidance on deployment security, cloud environment selection and organizational security management measures.

Link: https://www.tc260.org.cn/portal/article/2/160310cea5f6411d92fd99a52a42424f

8. National Data Administration Solicits Public Comments on the Guidelines for Data Property Rights Registration (Trial)

On April 3, the Department of General Affairs of the National Data Administration issued an announcement to solicit public comments on the Guidelines for Data Property Rights Registration (Trial) (Draft for Public Comments), with the comment period from April 3, 2026 to April 19, 2026. The Guidelines aim to establish and improve the data property rights system, build a unified national data property rights registration system, and cultivate a national integrated data market. The draft for comments clarifies that data property rights refer to the property rights enjoyed by obligees over specific data, including data holding rights, data use rights, data operation rights, etc., and are applicable to the property rights registration of various forms of data such as data resources and data products within the territory of China.

Link: https://mp.weixin.qq.com/s/ObJb4KVPJ9Pi9QuLpF18DQ

9. State-owned Assets Supervision and Administration Commission of the State Council, Ministry of Transport and Other Departments Issued the Intelligent Shipping 2030 Action Plan

On March 30, the Ministry of Transport, together with the Ministry of Industry and Information Technology and three other departments, jointly issued the Intelligent Shipping 2030 Action Plan. The plan sets two-stage development goals: by 2027, realize the in-depth integration of artificial intelligence and shipping elements, build more than three comprehensive pilot areas for intelligent shipping, open up more than five pilot routes, create more than ten typical scenarios, and operate more than 100 intelligent ships; by 2030, fully master the core and key technologies, form a new model of collaborative development of technology, industry and governance, and reach the international advanced level. The action plan deploys 11 key tasks from four dimensions: technical equipment research, application pilot empowerment, infrastructure improvement and supervision and governance improvement, including strengthening the research on common technologies for intelligent shipping, promoting the large-scale application of mature AI technologies, promoting the digital upgrading of communication and navigation facilities, building an integrated support platform and improving the supervision and governance system. It will provide replicable and scalable experience models for the development of intelligent shipping by carrying out pilot demonstration projects in representative regions and routes across the country.

Link: https://xxgk.mot.gov.cn/jigou/haishi/202603/t20260330_4202678.html

10. National Information Technology Standardization Technical Committee Issued a Notice on Soliciting Comments on 22 National Standards including Data - Basic Terms (Draft for Comments)

On April 1, the National Information Technology Standardization Technical Committee issued a notice on soliciting comments on 22 national standards, including Data - Basic Terms, Implementation Guide for Public Data Resource Registration, Terminology for Citywide Digital Transformation, Quality Evaluation Specification for High-quality Data Sets, Evaluation of Data Service Capability - Part 3: Implementation Guide, Implementation and Evaluation Guide for Data Anonymization Circulation, Basic Requirements for Interconnection of Data Infrastructure, Identification Requirements for Data Infrastructure, etc.

Link: https://mp.weixin.qq.com/s/x-8NuwzWRgE_Q429h6qeOA

(II) Law Enforcement and Judicial Dynamics

1. National Intellectual Property Administration Issued the Risk Warning on Using OpenClaw and Other Agents to Draft Patent Application Documents

On April 1, the National Intellectual Property Administration issued the Risk Warning on Using OpenClaw and Other Agents to Draft Patent Application Documents on its official WeChat account, pointing out three security risks that may arise when using OpenClaw to draft patent application documents. First, the risk of technical information leakage. Agents such as OpenClaw have hidden dangers such as excessive permissions, security vulnerabilities, and plugin poisoning. Using them to draft application documents may easily lead to the leakage of core information such as technical disclosure documents. Second, the risk of substantial defects. When using such agents to draft application documents, hallucinations may occur, resulting in problems such as contradictory content logic and unclear expression of technical features in the application documents, which will make it impossible to obtain protection. Third, the risk of dishonest application. A patent application formed by凭空 generation, random fabrication, and content piecing together through agents constitutes a dishonest patent application act that violates the principle of good faith. If the number reaches a certain level, the applicant will face administrative penalties such as warnings and fines; the agency and patent attorneys will face administrative penalties such as revocation of practice licenses and cancellation of agent qualification certificates, and will be included in the list of serious violations and dishonesty in serious cases.

Link: https://mp.weixin.qq.com/s/6ZqIze99Dx9veViu_HhnvQ

On April 2, the Inspection and Testing Center for Computer Information System Security Products of the Ministry of Public Security detected 37 mobile applications that illegally collected and used personal information in violation of laws and regulations. Among them, 21 mobile applications failed to disclose the rules for collection and use; 7 mobile applications failed to list the purpose, method and scope of collection and use of personal information one by one; 5 mobile applications failed to simultaneously inform users of the purpose when applying for opening permissions that can collect personal information; 5 mobile applications started to collect personal information before obtaining the user's consent; 7 mobile applications actually collected personal information beyond the scope of user authorization; 4 mobile applications required users to open permissions that can collect personal information not required for the current function in advance; 2 mobile applications failed to provide users with a specific way to delete their personal information; 4 mobile applications failed to provide users with the way and method to cancel their accounts; 1 mobile application required more types of personal information than those collected during registration, use and other service links in the process of identity verification for account cancellation.

Link: https://mp.weixin.qq.com/s/jR6AqawzJNpVliCiocJLnQ

3. 71 Mobile Applications Violating Laws and Regulations on the Collection and Use of Personal Information Detected by the National Computer Virus Emergency Response Center

On March 31, the National Computer Virus Emergency Response Center reported that 71 mobile applications were found to have illegally collected and used personal information in violation of laws and regulations. Among them, 17 mobile applications failed to prominently prompt the privacy policy, defaulted to consent, made the policy difficult to view, and failed to fully inform the necessary matters related to personal information processing; 45 mobile applications failed to list the relevant contents in the privacy policy one by one; 16 mobile applications illegally provided the personal information they processed to other personal information processors; 1 mobile application started to collect personal information or open permissions that can collect personal information without obtaining the user's consent; 4 mobile applications had non-compliant functions for correcting and deleting personal information and canceling user accounts; 1 mobile application's personal information processor failed to establish a convenient application acceptance and processing mechanism for individuals to exercise their rights; 24 mobile applications had improper ways and channels for withdrawing consent to collect personal information; 1 mobile application failed to provide options not targeted at personal characteristics or a convenient way for individuals to refuse when conducting information push and commercial marketing to individuals through automated decision-making; 2 mobile applications had improper processing of sensitive personal information; 6 mobile applications had improper protection of minors' information; 18 mobile applications failed to take corresponding security technical measures such as encryption and de-identification; 6 mobile applications had improper use of face recognition technology; 4 mobile applications had no privacy policy.

Link: https://mp.weixin.qq.com/s/aI36S5HCT7fUFiQBV6rndw

4. Cyberspace Administration of China, Ministry of Industry and Information Technology and Ministry of Public Security Launched the 2026 Series of Special Actions for Personal Information Protection

On April 2, the Cyberspace Administration of China, the Ministry of Industry and Information Technology and the Ministry of Public Security issued the Announcement on Launching the 2026 Series of Special Actions for Personal Information Protection. This special action targets the following seven aspects: special governance of illegal and irregular collection and use of personal information by Apps and SDKs; special governance of illegal and irregular collection and use of personal information in the field of Internet advertising; special governance of illegal and irregular collection and use of personal information in the field of education; special governance of illegal and irregular collection and use of personal information in the field of transportation; special governance of illegal and irregular collection and use of personal information in the field of health; special governance of illegal and irregular collection and use of personal information in the financial field; special crackdown and governance of criminal cases related to personal information.

Link: https://www.cac.gov.cn/2026-04/02/c_1776867645836849.htm

(III) Industry Trends

1. Office of the Privacy Commissioner for Personal Data of Hong Kong Held the "Artificial Intelligence Security and Cybersecurity Enterprise Summit"

On March 31, the Office of the Privacy Commissioner for Personal Data (PCPD) of Hong Kong held the "Artificial Intelligence Security and Cybersecurity Enterprise Summit" at the Hong Kong Convention and Exhibition Centre. This summit focused on the privacy challenges and cybersecurity defense pain points faced by enterprises in massive data processing under the background of the rapid popularization of generative artificial intelligence technology. Through real cross-border cases, the conference discussed how to build a robust data defense line that can not only release the potential of AI data, but also ensure cybersecurity and compliance.

Link: https://www.hkcd.com.hk/hkcdweb/content/2026/04/01/content_8747995.html

2. 2026 (3rd) Conference on Integration and Application Innovation of Data Elements Held in Beijing

On March 31, the 2026 (3rd) Conference on Integration and Application Innovation of Data Elements was held in Haidian, Beijing. Hosted by the China Electronics Chamber of Commerce, the conference was attended by more than 200 guests from government departments, scientific research institutions, universities and enterprises, with the theme of "Digital Intelligence Empowers全域 Integration", to jointly explore the path of integration and innovation of data elements. Wang Ning, President of the China Electronics Chamber of Commerce, said in his speech that 2026 is China's "Year of Releasing the Value of Data Elements" and also the final year of the three-year action plan of "Data Element ×". China's data element market has entered a critical period of large-scale development and value release. He pointed out that the industry has two major pain points: ownership and circulation, security and application, and put forward four development suggestions: improving the system, smoothing circulation, deepening innovation, and adhering to security.

Link: https://www.cecc.org.cn/wearable/news/view/id/573711

II. International Governance Developments

(I) Policy and Legislative Updates

1. Oregon Advances Special Regulation on Artificial Intelligence Companions

On March 31, the Oregon Legislative Information System showed that Senate Bill 1546 (SB 1546) has entered the Governor Signed status. The bill establishes special regulatory rules around "artificial intelligence companions", which are defined as systems that use artificial intelligence, generative artificial intelligence, or algorithms that recognize emotions from input, and simulate sustained human-like platonic, intimate, romantic or companion relationships with users by retaining the user's past interactions, conversations and preference information.

The bill requires operators to fulfill a clear obligation of identity disclosure. If a reasonable person in the corresponding situation would believe that they are interacting with a natural person, the operator shall clearly prompt the user that the content they are exposed to is artificially generated. For minors, the bill sets stricter restrictions, prohibiting operators from enabling the system to simulate emotional dependence, romantic interest, sexual innuendo, or carry out romantic role-play between adults and minors when they know or have reason to believe that the user is a minor. For high-risk situations such as suicide and self-harm, the bill also requires operators to establish a crisis identification and referral mechanism, and stipulates an annual reporting obligation; individuals who suffer identifiable losses may file a civil lawsuit to claim actual damages or statutory damages of $1,000 per violation, and apply for injunctive relief.

Link: https://olis.oregonlegislature.gov/liz/2026R1/Measures/GovernorSignedBills

2. U.S. Department of Health and Human Services Restructures Health Technology Leadership Framework to Promote Data Circulation and AI-enabled Healthcare

On March 31, the U.S. Department of Health and Human Services (HHS) announced the revocation of the 2024 medical technology restructuring arrangement, restored the department's unified technology leadership model, and clarified the division of responsibilities of various institutions. This adjustment reassigns the three functions of Chief Technology Officer, Chief Artificial Intelligence Officer and Chief Data Officer from the Office of the National Coordinator for Health Information Technology (ONC) back to the Office of the Chief Information Officer (OCIO) for management. OCIO will operate in an integrated manner around the three core functions of strategic technology innovation, trusted AI application, and enterprise data governance, and strengthen the statutory responsibilities of enterprise-level IT, cybersecurity and data operation. ONC will focus on the interoperability of national health information technology, data circulation and policy standard formulation, and continue to report directly to the HHS Secretary. HHS stated that this structural adjustment can integrate technical infrastructure and security capabilities, promote the efficient circulation of medical data, deepen the application of AI technology in clinical medical treatment, help build a safe, inclusive and efficient AI-enabled medical system, improve the quality of medical services and reduce the medical costs of the public.

Link: https://www.hhs.gov/press-room/hhs-health-tech-leadership-deliver-data-liquidity-affordability-ai-enabled-health-care-system.html

3. California Strengthens Safeguards for Government Artificial Intelligence Procurement and Use

On March 30, Gavin Newsom, Governor of California, signed Executive Order N-5-26, requiring the state government to further improve the security, privacy and rights protection mechanisms in the procurement and use of artificial intelligence. According to the press release from the Governor's Office and the full text of the executive order, the order mainly makes arrangements in four aspects: state government procurement review, contract liability, internal use specifications and content transparent identification.

First, strengthening procurement access review. The executive order requires the California Department of General Services and the California Department of Technology to put forward new certification recommendations within 120 days, requiring enterprises intending to cooperate with the state government to explain their policies and safeguard measures to prevent the technology from being used for the dissemination of illegal content, harmful prejudice, and infringement of citizens' rights and civil liberties. Second, researching and improving contractor liability clauses. The order requires state government operating agencies to research and adjust contract liability to avoid the state government signing contracts with entities that have been judicially found to have illegally weakened privacy or civil liberties. Third, standardizing the internal use of the state government. The order proposes to provide employees with vetted generative artificial intelligence tools under appropriate privacy and cybersecurity safeguards, and update the state digital strategy to promote the use of artificial intelligence to improve the accessibility, transparency and accountability of public services. Fourth, refining data and content governance tools. The order requires the release of a data minimization toolkit to provide best practices, templates, special contract terms and project review checklists for departments involved in sensitive data collection; at the same time, it requires the state technology department to issue best practice guidelines for watermark identification of images and videos generated or significantly tampered with by artificial intelligence within 120 days.

Link: https://www.gov.ca.gov/2026/03/30/as-trump-rolls-back-protections-governor-newsom-signs-first-of-its-kind-executive-order-to-strengthen-ai-protections-and-responsible-use/

4. European Commission Updates Official FAQ on the Artificial Intelligence Act

On April 2, the European Commission updated the official FAQ on the definition of agents in the Artificial Intelligence Act (AI Act) and the revision of the Artificial Intelligence Act based on the consultations received during the AI Pact webinar and the questions submitted by stakeholders. The FAQ states that an agent contains at least one general artificial intelligence model, and because it usually has some form of interface, the agent falls within the scope of artificial intelligence systems. This means that the Artificial Intelligence Act also applies to agents. Among the relevant rules for artificial intelligence systems, the provisions of the Artificial Intelligence Act on prohibited harmful manipulation and exploitation of vulnerabilities (Article 5 (1) (a) and (b)) are particularly important. To comply with these provisions, safeguards must be set in the design and development process of agents to avoid the implementation of prohibited acts that may cause significant harm. From August 2, 2026, if an agent is classified as a high-risk artificial intelligence system, it must also comply with additional requirements to ensure its safety and trustworthiness in the intended use scenarios. For the general artificial intelligence models that form the basis of agents, the degree of autonomy of the model, tool use capability and other factors are the key to determining whether it is a model with systemic risk. In addition, providers of general artificial intelligence models with systemic risk must fulfill risk management obligations, including considering the autonomous capability of the model and its agent-like application scenarios.

Link: https://ai-act-service-desk.ec.europa.eu/en/faq

5. South Africa to Introduce Data Specifications for Access Control Venues Based on the Protection of Personal Information Act

According to a BusinessTech report on April 5, South Africa will introduce data specifications for access control venues based on the Protection of Personal Information Act (POPIA), which will strictly restrict excessive collection, random transcription or unsafe storage of visitor information by estates and residential complexes, including the scanning of driver's licenses and identity documents. The new regulations require that only necessary identity and access information can be collected, standardize the use of technologies such as license plate recognition and face recognition, and clarify that the responsibility for data security shall be borne by the residential complex and the owner's institution. Violations may face investigation, fines and even criminal liability.

Link: https://businesstech.co.za/news/property/855647/warning-for-estates-and-complexes-that-scan-visitor-car-and-drivers-licences-in-south-africa/

(II) Law Enforcement and Judicial Updates

1. Australian Government Signed a New Agreement on Artificial Intelligence Cooperation with Anthropic

On April 1, the Australian government signed a new memorandum of understanding on cooperation with Anthropic to further advance the National Artificial Intelligence Plan launched at the end of 2025. The memorandum states that Anthropic will promote artificial intelligence security research, monitor the productivity benefits of artificial intelligence and its broader economic and social impacts, predict the infrastructure needs for the development and deployment of cutting-edge artificial intelligence and develop solutions, support scientific research, education and skills training projects including small and medium-sized enterprises, and take measures to benefit local researchers. The memorandum of understanding establishes an action framework for cooperation, which will track the progress of artificial intelligence, promote security, support supply chain security, and promote Australia's R&D and innovation fields.

Link: https://www.minister.industry.gov.au/ministers/charlton/media-releases/new-agreement-ai-collaboration-anthropic

2. Swiss Finance Minister Files Defamation Lawsuit Over Post Created by Grok

According to a Reuters report on April 1, Karin Keller-Sutter, Swiss Finance Minister, filed a criminal complaint for defamation and insult over an obscene post generated by Grok and posted by an X user. Previously, an X user entered instructions to make Grok launch sarcastic insults against Keller-Sutter. The post generated by Grok used abusive language against her and her work, and then asked the user if they wanted more extreme content or to target other people. In response, a spokesperson for the Ministry of Finance stated that the technology may be used for crimes, and the prosecutor should review whether the relevant parties of X provided Grok with knowledge, and whether the relevant parties violated the duty of care.

Link: https://www.reuters.com/world/swiss-finance-minister-sues-defamation-over-grok-created-post-2026-04-01/

3. U.S. Government Appeals Ruling That Blocked Military Action Against Anthropic Over Artificial Intelligence Dispute

According to an Associated Press report on April 2, the U.S. government filed an appeal against a ruling that blocked the military from taking action against Anthropic over an artificial intelligence dispute. Previously, the U.S. military identified Anthropic as a supply chain risk, and Anthropic filed a lawsuit. The U.S. District Court for the Northern District of California issued an interim ruling in favor of Anthropic, blocking the instruction to mark Anthropic as a supply chain risk and requiring all federal agencies to stop using Anthropic's chatbot Claude. In response, the government has filed an appeal, and the case will be handled by the Ninth Circuit Court of Appeals.

Link: https://apnews.com/article/ai-pentagon-anthropic-claude-hegseth-trump-b24b0197f87658d0e8a60790a4c969a2

4. U.S. Federal Trade Commission Takes Enforcement Action Against Match Group and OkCupid

On March 30, the U.S. Federal Trade Commission (FTC) announced that it would take enforcement action against the dating platform OkCupid and its affiliated party Match Group Americas. The FTC filed a lawsuit in federal court and simultaneously disclosed the proposed settlement content, alleging that OkCupid violated its privacy commitments by disclosing users' personal information, including user photos and geographic location information, to third parties with no business relationship. The FTC stated that the platform was accused of allowing third parties to access the personal data of millions of users, and the relevant third parties allegedly obtained nearly 3 million user photos. According to the proposed settlement arrangement, OkCupid and Match will be permanently prohibited from making false or misleading statements about the collection, use, disclosure, deletion, protection methods of their personal information, as well as privacy control functions. The core of the case is not the abstract data compliance slogan, but whether the platform's external privacy commitments are consistent with the actual data sharing behavior, reflecting that the U.S. regulatory authorities continue to take the authenticity of privacy statements, the transparency of third-party data sharing, and the actual control measures of the platform as the key law enforcement direction.

Link: https://www.ftc.gov/news-events/news/press-releases/2026/03/ftc-takes-action-against-match-okcupid-deceiving-users-sharing-personal-data-third-party?utm_source=.com

(III) International Cooperation Dynamics

1. World Data Organization Established to Launch Global Data Governance Cooperation Framework

According to a Xinhua News Agency report on March 31, the World Data Organization was established and officially put into operation in Beijing. The organization is positioned as the first professional international organization focusing on data development and governance practices, with its tenets including "bridging the data divide, releasing data value, and prospering the digital economy", providing a cooperation platform for promoting the coordination of global data rules and AI data governance. Its members are composed of enterprises, institutions and experts from many countries and regions, and will become an important institution to promote international cooperation among countries in data sharing, algorithm transparency, cross-border data flow and AI data ethics in the future.

Link: https://www.xinhuanet.com/tech/20260331/8e98bb1c209047148138a8425acb71ba/c.html

2. Indonesia and Japan Reach Intent for Artificial Intelligence Cooperation

On March 31, Japanese Prime Minister Sanae Takaichi and Indonesian President Prabowo held talks in Tokyo and reached an intent to strengthen cooperation between the two countries in the fields of artificial intelligence talent training and application capacity building. This intentional cooperation includes supporting Indonesia in developing AI-related talent teams, promoting Japan to help improve Indonesia's AI education and R&D ecosystem, and may further expand collaborative innovation in the fields of maritime security and digital infrastructure in the future. The report also pointed out that the two countries regard such cooperation as an important part of promoting scientific and technological cooperation in the "Free and Open Indo-Pacific".

Link: https://setkab.go.id/en/indonesia-japan-ready-to-become-pillars-of-global-stability/

3. Organisation for Economic Co-operation and Development Holds International Conference on "AI in Work, Innovation, Productivity and Skills"

From March 30 to April 1, the Organisation for Economic Co-operation and Development (OECD) held the 2026 International Conference on "AI in Work, Innovation, Productivity and Skills". The conference focused on the impact of artificial intelligence on employment, skills, productivity, innovation and competition, with participants covering policy makers, enterprises, academia and civil society. In terms of process, this is one of the international dialogue mechanisms continuously promoted by the OECD around AI and labor market governance. This conference was held online and set up multiple special discussions. Its main content was to exchange views on topics such as "agent-based AI", "algorithmic management", "the impact of AI on productivity and innovation", and "how AI supports neurodiverse learners", and discuss how policies should respond to the rapidly expanding AI applications. The reason why this development is noteworthy is that it reflects that international organizations are further advancing the discussion on AI governance from purely technical and risk issues to cross-sectoral issues such as labor, education, industrial competitiveness and inclusive growth, with a strong color of international policy coordination.

Link: https://www.oecd.org/en/events/2026/03/2026-international-conference-on-ai-in-work-innovation-productivity-and-skills.html

4. U.S. and Italy Clarify Deepening of Artificial Intelligence Cooperation in the Joint Commission on Scientific and Technological Cooperation

On April 2, the U.S. Department of State released the Joint Statement of the 15th U.S.-Italy Joint Commission Meeting on Science and Technology Cooperation, stating that the United States and Italy committed to further strengthening cooperation in the fields of artificial intelligence, high-performance computing, quantum science, fusion and fission energy, biotechnology and others under the framework of scientific and technological cooperation on the same day. The meeting was hosted by the United States with Italy participating as the cooperating party, and the two sides formed a joint statement on the key future cooperation directions under the existing scientific and technological cooperation mechanism. Its main content is not the signing of a single AI project, but the formal inclusion of AI into the key topics of bilateral scientific and technological cooperation, reflecting that the two countries hope to form closer coordination in cutting-edge technology governance and innovation capacity building. This shows that AI has become an important part of bilateral scientific and technological diplomacy and strategic cooperation, not just an industrial or regulatory issue. At the same time, AI is listed alongside high-performance computing and quantum technology, which also reflects its fundamental position in national scientific and technological competition and cooperation.

Link: https://www.state.gov/releases/office-of-the-spokesperson/2026/04/joint-statement-of-the-15th-u-s-italy-joint-commission-meeting-on-science-and-technology-cooperation/

5. World Economic Forum Calls for Inclusion of Artificial Intelligence Computing Power in National Critical Infrastructure

On April 2, the World Economic Forum (WEF) released a report, which comprehensively analyzed the profound impact of artificial intelligence computing infrastructure on the global economic operation model, and called for it to be regarded as critical public infrastructure. The report highlighted that driven by the deployment of AI accelerators and ultra-large-scale computing, the electricity demand of U.S. data centers is expected to soar to 6.7%-12% of the total social power consumption by 2028; at the same time, as AI inference workload is expected to exceed model training by 2030, the marginal cost of computing power is dropping sharply. The WEF recommended that governments of all countries make policy adjustments at the legal, urban planning and national security architecture levels as soon as possible, and manage AI computing power as a core infrastructure for maintaining the modern economy and public services in an integrated manner.

Link: https://www.weforum.org/stories/2026/04/ai-infrastructure-critical-infrastructure/

(IV) Research Developments

1. United Nations Educational, Scientific and Cultural Organization and Thomson Reuters Foundation Released Report on Corporate Artificial Intelligence Governance

On March 31, the United Nations Educational, Scientific and Cultural Organization (UNESCO) and the Thomson Reuters Foundation jointly released Responsible AI in Practice: 2025 Global Insights from the AI Company Data Initiative. According to the official UNESCO website, the report is based on the analysis of public information from 3,000 enterprises, focusing on the actual status of enterprises in terms of artificial intelligence adoption, governance framework, supervision mechanism and responsibility implementation.

The report shows that about 44% of companies stated that they have an artificial intelligence strategy, but only about 10% of companies have publicly committed to following a certain artificial intelligence governance framework; only 12% of companies have policies to ensure that artificial intelligence systems retain human oversight; the disclosure ratio is also relatively limited in terms of environmental and human rights impact assessment. The report believes that the main problem faced by current corporate artificial intelligence governance is no longer whether they are aware of the ethical and governance risks of artificial intelligence, but whether they can transform governance principles into verifiable, disclosable and accountable daily operation mechanisms within the organization.

Link: https://www.unesco.org/en/articles/pioneering-report-thomson-reuters-foundation-and-unesco-sheds-light-way-3000-companies-approach-ai

2. Federal Reserve Released Monitoring Report on Trends in Artificial Intelligence Adoption in the U.S. Economy

On April 3, the Federal Reserve released a research report that systematically monitors the adoption of artificial intelligence (AI) in the U.S. economy. The report overall integrates a number of public market survey data, aiming to quantitatively analyze the penetration rate of AI applications in U.S. enterprises. Key data shows that by the end of 2025, about 18% of U.S. enterprises had adopted AI technology; in terms of employment weight, up to 78% of the U.S. labor force worked in enterprises that had adopted AI, and 54% of the labor force was in enterprises that applied large language models (LLMs). The report assesses that AI deployment by large enterprises will significantly accelerate the diffusion of the technology among the labor force, and it is expected that the U.S. will usher in a new wave of enterprise-level AI adoption in the first half of 2026.

Link: https://www.federalreserve.gov/econres/notes/feds-notes/monitoring-ai-adoption-in-the-u-s-economy-20260403.html

(V) Industry Dynamics

1. Digital Economy Cooperation and Innovation Forum of the 2026 Global Digital Economy Conference Held in Geneva

On March 31, the Digital Economy Cooperation and Innovation Forum of the 2026 Global Digital Economy Conference was held at the Palais des Nations in Geneva, Switzerland. During the event, the Digital Economy City Alliance (DEC40) announced the official launch of its Geneva office, and 11 new Chinese and Swiss entity enterprises joined the alliance. In addition, the forum jointly released the global digital economy beacon cases, and launched an international innovation competition focusing on artificial intelligence generated content (AIGC).

Link: https://english.news.cn/europe/20260401/0ad3b8345b004e798a49b82b4e809e83/c.html

ꄴPrevious： null

ꄲNext： null