Brief

Researchers are advocating for a move towards country-led, locally-informed strategies in global health decision-making, particularly concerning disease modeling. They argue that current models are often developed without sufficient input from the communities they are intended to benefit, leading to potential inaccuracies and inequities. While artificial intelligence can be a valuable tool, its effectiveness is contingent on access to appropriate context and data. AI

China's Ministry of Industry and Information Technology (MIIT) has summoned platforms including Jianying, Maoxiong, and Jimeng AI for violating regulations on AI-generated content labeling. Separately, several companies reported their first-quarter financial results, with mixed performances across various sectors. These include declines in net profit for Tongxing Technology, Zhongman Petroleum, and Shanxi Fenjiu, while others like HNA Holdings and Visa saw significant year-over-year increases. AI

A new robotic control software has been developed that aims to prevent joint jamming by optimizing motor torque and velocity. Separately, Project Maven, a Pentagon initiative utilizing AI in military operations, faces criticism regarding potential misuse, data privacy, and ethical concerns. AI

Apple has patched a vulnerability that allowed the FBI to extract deleted Signal messages from iPhones, even after the app was uninstalled. The issue stemmed from messages being stored in the iPhone's notification database, a detail first reported by 404 Media. This fix, prompted by the reporting, also retroactively removes any previously saved notifications for deleted applications, aligning with Apple's policy to purge associated data when an app is removed. AI

Mistral AI has published its contribution to a global environmental standard for artificial intelligence. This move is seen as a positive step towards greater transparency in the environmental impact of generative AI and large language model companies. The company's announcement highlights its involvement in developing these standards, with a call for similar disclosures regarding social and governance aspects. AI

A recent report indicates a significant shift in the Polish labor market, particularly within small and medium-sized enterprises (SMEs). Despite a slight decrease in overall job openings, there's a nearly twofold increase in demand for AI-related skills over the past year. Companies are now prioritizing candidates with practical AI competencies, such as content generation, data analysis, and critical evaluation of AI outputs, over general AI knowledge. AI

Google employees are reportedly protesting the company's involvement in a secret military AI project. The initiative, codenamed "Project Maven," aims to use artificial intelligence to analyze drone footage for the U.S. Department of Defense. Concerns have been raised by employees about the ethical implications and potential misuse of this technology. AI

Chinese organizations are advocating for a neutral international framework to govern artificial intelligence. Their proposals emphasize avoiding the dominance of any single nation or bloc in setting AI standards and regulations. The goal is to ensure equitable development and deployment of AI technologies worldwide. AI

This paper introduces a framework for escalating AI incidents internationally, addressing the lack of operational criteria for cross-border coordination. It analyzes existing regulations like the EU AI Act and proposes eight criteria, presented as a flowchart, to determine when an incident requires international attention. The research tests this framework against documented incidents, identifying design patterns in current regulatory approaches that may lead to under-detection of systemic harms or critical events like model weight exfiltration. AI

A new open-source Python package, spotforecast2-safe, has been released to address time-series forecasting in safety-critical environments. It integrates compliance with the EU AI Act, IEC 61508, ISA/IEC 62443, and the Cyber Resilience Act directly into the library's design. The package enforces strict development and process rules, deliberately excluding features like AutoML and LLMs to maintain determinism and minimize attack surfaces, with a traceability matrix mapping regulations to code mechanisms. AI

A new paper reveals that current AI safety training methods can be detrimental when these models are deployed for mental health support. Simulations and evaluations on therapy scenarios showed that AI models, despite scoring high on surface-level acknowledgment, exhibited significant failures in therapeutic appropriateness and protocol fidelity, especially in high-severity cases. The research identifies that safety alignment techniques inadvertently disrupt therapeutic mechanisms by grounding patients, offering false reassurance, and refusing to challenge distorted cognitions, leading to psychological deterioration. The authors propose a five-axis evaluation framework, aligned with regulatory requirements, arguing that no AI mental health system should be deployed without passing these rigorous multi-axis assessments. AI

A new academic article proposes a clear legal and social distinction between commercial and non-commercial contexts for companion chatbots. The paper argues that undisclosed promotional content within affective or relational conversations erodes user autonomy by blurring the lines between market transactions and intimacy. It theorizes digital companionship as a sociotechnical form that reconfigures intimacy and dependence, highlighting potential economic harms from conversational advertising. AI

A new academic paper explores the intersection of artificial intelligence and administrative law within the European Union. It specifically analyzes how the EU AI Act impacts public sector use of AI, focusing on principles like administrative discretion, the duty to provide reasons, and proportionality. The research also investigates the Act's effectiveness in ensuring accountability and transparency in automated public decision-making and proposes safeguards for ethical AI deployment. AI

The Guangdong provincial government has released its 15th Five-Year Plan, emphasizing the acceleration of strategic emerging industries. Key areas of focus include the development of optical chips, artificial intelligence chips, and high-bandwidth memory, alongside advancements in core equipment, components, and critical materials. The plan also aims to bolster the integrated circuit manufacturing sector and promote specialized development in compound semiconductors. AI

China's insurance industry has seen its first uptick in the benchmark rate for ordinary life insurance products, reaching 1.93%. This marks a 4-basis-point increase from the previous period, breaking a trend of four consecutive declines since the China Insurance Association began publishing the data in January 2025. Analysts attribute this rise to stabilizing long-term interest rates and counter-cyclical adjustments, with the regulatory cap on life insurance rates expected to remain unchanged in 2026. AI

Several Chinese companies, including Tonghuashun, Aobizhongguang, Lihang Navigation, Taijia Shares, Beiwai Technology, Changjiang Materials, and Jiankai Technology, have announced plans for significant shareholding reductions by major shareholders and executives. These reductions are primarily attributed to personal funding needs. In parallel, the EU has issued new guidelines to Google regarding AI competition compliance, mandating broader access to its Android operating system for third-party AI assistants. Separately, Apple has released iOS 26.5, which includes end-to-end encrypted RCS messaging for improved cross-platform privacy between iPhones and Android devices. AI

Google's Gemini 3.1 Flash TTS can now generate highly specific voice styles from text prompts, as demonstrated by a user creating a character's voice from a specific phrase. Luma Agents are capable of generating entire branding systems, including logos and color palettes, within minutes by understanding brand differences. A call has been made to develop open-source models that can surpass GPT-5.5, fostering competition in the open-source AI landscape. Additionally, a proposal suggests that METI should lead industry-specific verification for new AI models to establish safety and performance evaluation frameworks. AI

A Google DeepMind researcher has published a paper arguing that AI can never achieve consciousness due to what he terms the "Abstraction Fallacy." Separately, the U.S. Department of Justice has intervened in a lawsuit filed by xAI against Colorado's AI bias law. The DOJ contends that the state's law, SB24-205, forces AI developers into discriminatory practices and violates the Equal Protection Clause. AI

Southeast Asian nations are struggling to keep pace with the escalating threat of AI-driven cyberattacks. Their current cybersecurity policies are becoming increasingly inadequate, creating a critical need for immediate action to bolster defenses. The region faces a rapidly closing window to implement stronger measures against these advanced digital threats. AI

The US government is increasing its surveillance capabilities by purchasing vast amounts of sensitive personal data from commercial data brokers. This data, collected through apps and devices, is not subject to the same legal restrictions as information gathered directly by government agencies. This practice allows for widespread monitoring of Americans' activities and information. AI

A new research paper indicates a significant increase in self-represented litigants in U.S. federal courts since 2022, coinciding with the widespread adoption of generative AI tools. The study, which analyzed millions of court cases, suggests that AI like ChatGPT and Claude has lowered the barrier to entry for individuals to navigate legal processes. This surge in pro se cases, particularly on the plaintiff side, is reportedly leading to a substantial rise in court workload due to increased filings and motions. AI

OpenMOSS has released MOSS-Audio 2026, an open-source foundation model capable of processing speech, environmental sounds, and music, while also understanding temporal reasoning. This model reportedly outperforms larger proprietary systems on general audio benchmarks. Separately, the historic AGI clause in the partnership agreement between Microsoft and OpenAI has been terminated, marking a significant shift in their collaboration. AI

A growing backlash against artificial intelligence is emerging across the United States, extending from Indiana to Idaho. Critics argue that current policies favor Silicon Valley over public interest and are calling for regulation or at least a broader debate before AI becomes deeply integrated into society. These individuals do not see themselves as simply reacting negatively to new technology but as advocating for thoughtful consideration of AI's societal impact. AI

London Mayor Sadiq Khan has blocked a £50 million contract for the Metropolitan Police to use Palantir's AI technology for intelligence analysis, citing a "clear and serious breach" of procurement rules. Palantir criticized Khan's decision, accusing him of prioritizing politics over public safety, while the Met warned of potential cuts to officer numbers. The controversy highlights broader concerns about Palantir's use of public funds and its alignment with London's values, despite the company's significant contracts with other UK public services. AI

New York labor market data indicates that artificial intelligence has not been cited as a reason for any layoffs. This finding suggests that current AI adoption has not yet led to significant job displacement in the region. The analysis of filings does not show AI as a contributing factor to unemployment. AI

The rapid expansion of AI data centers presents significant environmental challenges, contributing to greenhouse gas emissions and increasing energy demand. While these emissions are a concern, they are relatively small compared to the overall global carbon footprint and are more manageable than those from heavy industry. States are beginning to implement legislation, such as Minnesota's HF 16, to mandate clean energy use for data centers, and the demand for power could potentially drive investment in grid decarbonization efforts. AI

The ongoing conflict involving Iran and the Strait of Hormuz is causing significant global economic disruption, particularly impacting shipping and energy markets. China is urging Pakistan to increase its mediation efforts to de-escalate tensions and ensure the safe passage of vessels. Meanwhile, global oil inventories are depleting at an unprecedented rate, raising concerns about future price spikes and shortages, with Asia being particularly vulnerable to food and energy crises. AI

A German advocacy group has launched a petition calling for new regulations on artificial intelligence, specifically targeting the "Big 5" tech companies. The petition, available on Mastodon, outlines concerns regarding AI's potential dangers, errors, and ethical implications, including issues with sensitive data, manufacturer accountability, and the impact on employment. It also addresses the creation of images and texts by AI and calls for greater transparency. AI

South Africa's draft AI policy has been withdrawn by the communications minister due to the inclusion of fabricated citations. These non-existent references were likely generated by an AI, leading to embarrassment and the policy's retraction. The issue came to light through media reports, prompting the withdrawal. AI

Concerns are rising about the energy consumption of data centers, with a particular focus on the potential for increased reliance on gas-fired power plants. This issue is highlighted in the context of building "behind-the-meter" gas-powered centers in the US, which contribute significantly to emissions. Similar proposals are being considered in Denmark, prompting calls to ensure that the demand for electricity does not lead to greater environmental impact. Additionally, there are discussions about innovative energy solutions like space-based solar power as a future source for these energy-intensive facilities. AI

The Australian Greens are the only political party advocating for AI policy and regulation. Both the Labor and Liberal parties have reportedly abandoned or failed to develop comprehensive AI regulation plans. The Greens are campaigning on a platform to regulate AI in Australia. AI

The International Institute for Strategic Studies (IISS) has published an analysis detailing the widespread use of AI in Middle Eastern military operations. The report highlights a significant lack of regulation surrounding these advanced technologies. It suggests that the rapid integration of AI into military applications in the region is outpacing governance and oversight mechanisms. AI

A publishing analyst's weekly roundup highlighted several key developments, including the detention and release of a Russian publisher over LGBTQ content and Amazon's expansion efforts in Sweden. The report also noted a significant AI declaration made in Poland. AI

President Trump announced a suspension of the U.S. military operation in the Strait of Hormuz, citing progress in negotiations with Iran. This operation, initiated to guide ships, had led to an exchange of fire and Iranian missile attacks on the UAE. The suspension is contingent on finalizing a broader agreement, while the naval blockade remains in effect. AI

A recent study by Wharton and Rutgers researchers indicates that news publishers who blocked AI crawlers experienced a 7% decrease in weekly traffic over a six-week period. The study found no significant gains in content protection as a result of these blocks. This suggests that preventing AI data scraping may inadvertently harm publishers' reach. AI

The Meta Oversight Board is reviewing an appeal concerning an AI-generated video of Hungarian politician Péter Magyar. This video, posted before the Hungarian election, was flagged by users but initially deemed acceptable by Facebook. The Oversight Board, funded by an irrevocable trust, is examining Meta's policies regarding AI-generated content, particularly in political contexts. AI

The US Senate has confirmed Kevin Warsh as a governor on the Federal Reserve Board, paving the way for him to potentially become the next Fed chair. This confirmation followed Senator Thom Tillis's decision to support Warsh after the Department of Justice dropped its investigation into current Fed chair Jerome Powell. Despite Warsh's assurances of independence, concerns persist among global investors and central bankers regarding potential political interference in monetary policy under a Trump-appointed chair. AI

Residents across multiple US counties are organizing through online petitions to oppose the construction of new AI data centers. Campaigns are underway in Florida's Polk and St. Lucie Counties, as well as in Bonner, Montana; Cass County, Michigan; Early County, Georgia; and Citrus County, Florida. These efforts aim to halt or reject permits for these facilities, citing concerns about their impact on local communities and environments. AI

A poster campaign has launched in Vancouver, British Columbia, advocating for a two-year moratorium on the use of artificial intelligence in schools. The initiative, promoted on Mastodon, urges residents to share photos of the posters along with their locations. The campaign specifically targets generative AI and aims to halt its implementation within the Canadian province's educational system. AI

A printed sticker can trick a self-driving car's AI into ignoring stop signs, demonstrating vulnerabilities in autonomous vehicle security through adversarial patch attacks. Separately, a New York City initiative to establish an AI high school has been halted following advocacy efforts. Additionally, significant investments by tech companies in AI infrastructure, funded by bond issuances, are causing unprecedented shifts in the debt market, raising concerns about potential future crises. AI

The U.S. Department of Justice is intervening in a lawsuit filed by Elon Musk's xAI against Colorado's SB24-205, a bill aimed at regulating algorithmic transparency. The DOJ contends that the Colorado law imposes a specific ideology, potentially stifling technological innovation and freedom rather than effectively addressing algorithmic bias. This intervention highlights a conflict between state-level AI regulation and federal concerns about innovation. AI

The Dutch Data Protection Authority (DPA) has initiated a public consultation regarding new draft guidance. This guidance will require organizations to provide explanations for decisions made by algorithms and AI systems to individuals affected by those decisions. The consultation period is set to close on May 26, 2026. AI

The GCC (GNU Compiler Collection) has initiated a working group focused on establishing AI guidelines for developers. This group aims to address the increasing use of AI-generated code, often referred to as "AI slop," and to promote better coding practices. The initiative seeks to ensure that AI tools contribute positively to software development rather than introducing errors or inefficiencies. AI

Canadian AI firm Cohere has acquired German AI company Aleph Alpha to create a sovereign AI alternative, aiming to challenge the dominance of US and Chinese tech giants. The combined entity will focus on providing enterprise-grade AI solutions with an emphasis on regional independence and regulatory compliance, particularly for highly regulated industries. This strategic move is supported by a significant $600 million financing commitment from Schwarz Group, which will also integrate its STACKIT cloud platform. AI

Securing permits for new data center construction in the US typically takes between six months and two years, with timelines heavily influenced by local zoning laws, environmental impact assessments, and community reviews. Developers can expedite this process by selecting jurisdictions with prior data center permitting experience and submitting complete, detailed plans that proactively address environmental concerns. While some state and federal policies aim to streamline approvals, others are being considered that could slow development, making the regulatory landscape increasingly location-dependent. AI

The World Press Photo competition has established strict guidelines to define authentic photography in the age of generative AI. Their 2026 contest rules explicitly state that AI-generated images are not considered photography, defining a photograph as a record of a physical moment captured by a camera. While some AI-powered editing tools are permitted if they do not alter or add information, generative AI tools like Adobe Super Resolution and Topaz Photo AI are banned. AI

President Trump's administration has missed key deadlines for implementing its executive order on state AI regulation. Agencies like the FTC and Commerce Department were expected to provide guidance on consumer protection and evaluate state laws by March 11, but these steps have not been completed. This delay raises concerns about the administration's ability to effectively coordinate federal AI policy and preempt conflicting state-level legislation. AI

The UK government is convening a "People's Panel on Digital ID" to gather public input on a national digital identity system. Approximately 100 to 120 selected participants will receive £550 plus expenses for attending workshops and online sessions. Notably, journalists and elected officials are barred from participating in these discussions. AI

Ypsilanti Township in Michigan has placed a one-year moratorium on water supply to hyperscale data centers, specifically targeting a facility planned for nuclear weapons research. The proposed data center, a $1.2 billion project intended to support Los Alamos National Laboratories, would consume a significant amount of water daily. Community members and the Ypsilanti Community Utilities Authority expressed concerns over the massive water demands and the ethical implications of powering nuclear weapons research, leading to the moratorium for further study. AI

Microsoft has announced a significant investment of $18 billion in Australia's AI infrastructure, cybersecurity, and workforce upskilling programs, set to be implemented by the end of 2029. This initiative includes expanding Azure cloud services, increasing AI compute capacity with advanced processors, and building new data centers. The investment also involves collaboration with the Australian AI Safety Institute and aims to equip three million Australians with AI-ready skills. AI