hotai.newsAn archive of 20 AI news briefs collected on 2026-09-22. Check the original sources for details and updates.
This content still needs to be cross-checked against the original publication page.
GPU acceleration can speed up compute-intensive robotics workloads, but a fast CUDA kernel alone does not guarantee a fast ROS 2 graph. As messages move between…
Source links can be verified on the original siteRead original ↗The British AI data center developer depends on tech giants Microsoft and Anthropic for most of its revenue.
Source links can be verified on the original siteRead original ↗Certainly the industry is transforming, however, the people and organizations falling into the trap of no longer reading and writing code only do so at their peril.
Source links can be verified on the original siteRead original ↗An Academy for Unusually Ambitious Young Builders
Source links can be verified on the original siteRead original ↗The latest release brings AI agent capabilities to the ROS developer ecosystem while expanding open source physical AI libraries and deployment across the NVIDIA Jetson platform.
Source links can be verified on the original siteRead original ↗Meta has issued a patch for its Muse macOS app following the discovery of a zero-day vulnerability that could allow someone to take control of the AI agent.
Source links can be verified on the original siteRead original ↗AI can produce code. Organizations still have to produce software. Agentic development is changing how software gets made, but it hasn’t changed what it costs to be wrong. Six months ago, we began
Source links can be verified on the original siteRead original ↗With MiMo-V2.6-Pro, Xiaomi moves to the top of the openly available AI models and drastically undercuts the competition on price. What makes that possible is massive reinforcement learning that cost $2.62 million. But the success comes with a catch: Anthropic accuses the company of siphoning off training data from Claude.
Source links can be verified on the original siteRead original ↗39% of respondents believe AI is having a negative impact on society, up from 36% last month, reaching the highest level since surveys began on this issue in March. Only 11% believe AI brings positive impacts, while the remaining respondents expressed uncertainty or did not answer.
Source links can be verified on the original siteRead original ↗OpenAI says a new internal model solved more than 100 open math problems after just a month of training. Facing criticism from mathematicians, the company is backing an independent advisory group at the Institute for Advanced Study. But OpenAI has excluded its research pace from the group's advisory role.
Source links can be verified on the original siteRead original ↗School researchers reviewed nearly 200 studies on how artificial intelligence affects children and young people’s ability to learn and think. They found that we know very little about how AI use impacts younger school children.
Source links can be verified on the original siteRead original ↗An international AI science expert group spoke during the United Nations General Assembly in New York, calling for abandoning extreme rhetoric and distinguishing known facts from reasonable speculation. Experts emphasized that hyping fear is not conducive to scientific research, and the world needs a rational dialogue.
Source links can be verified on the original siteRead original ↗According to Reuters this evening (22nd), banks including NatWest and Bank of America have warned that letting AI agents shop online on behalf of consumers may increase the risks of fraud, scams, and data privacy breaches.
Source links can be verified on the original siteRead original ↗The model was developed by the Austrian Academy of Sciences together with multiple institutions, and its training corpus contains over 600 million Ancient Greek words. It can automatically complete damaged parts of texts and help improve the efficiency of ancient manuscript restoration; the final judgment is still completed by scholars.
Source links can be verified on the original siteRead original ↗A Futurism investigation found that Space Daily has an estimated 10 million monthly views. By traffic alone, it would have been a stunning success story. However, the site relies on fictional 'experts' that do not exist, then uses these seemingly authoritative identities to mass-produce AI spam content.
Source links can be verified on the original siteRead original ↗QwenNote Eva uses Qianwen Office as its core and DingTalk ecosystem as its base, supports proactive delivery via one-sentence commands, integrates AI transcription, meeting schedules, emotional companionship and 65W fast charging, claiming '1 Eva = 1 team'.
Source links can be verified on the original siteRead original ↗OpenAI and Anthropic urged Australia to reconsider its ban on the two companies using local creative content to train AI models, arguing that limited exemptions would both protect artists' rights and help attract more investment.
Source links can be verified on the original siteRead original ↗Anthropic is building its own biology lab to push AI-driven drug development beyond computer simulations.
Source links can be verified on the original siteRead original ↗According to The Information citing people familiar with the matter, before a recent series of cybersecurity incidents involving OpenAI technology occurred, and before industry insiders issued dense serious warnings, OpenAI had discussed with Anthropic a legally binding agreement under which the two sides would mutually conduct stress tests on each other's models.
Source links can be verified on the original siteRead original ↗Meta alone captured 81.3% market share with a shipment scale of 3.4 million units. China's AI glasses shipments surged 306% year-on-year in 2026H1.
Source links can be verified on the original siteRead original ↗