Technology News
545 articles
AI technology breakthroughs, research papers, patents, and technical innovations from leading AI startups.
Anthropic proposes CJS framework to assess AI jailbreak risks in 5 levels
Anthropic has published a draft of its Cyber Jailbreak Severity (CJS) framework, classifying jailbreak risks from CJS-0 to CJS-4. The system evaluates capability gain, breadth of capability gain, ease...
Expedia has published its AI and ML model principles based on processing billions of AI predictions,...
Expedia has published its AI and ML model principles based on processing billions of AI predictions, outlining a framework for building, deploying, and evolving AI systems for durable impact at scale....
Bespin Global publishes enterprise AI operations guide spotlighting HelpNow AI Foundry for data-governance-security integration
South Korean AI services and solutions company Bespin Global has released a Q2 2026 technical guide titled 'AI Technology Guide: Use It Right Now,' based on its internal newsletter Bespick. The guide...
Vercel CEO Guillermo Rauch argues for separating AI models from agents to optimize production price/performance
At Vercel's ShipNYC conference, CEO Guillermo Rauch articulated a strategic vision that directly addresses the core production challenge facing enterprise AI deployments: the separation of AI models f...
Tencent releases open-weight model Hy3 under permissive Apache license, outperforming GLM-5.2 on mos...
Tencent releases open-weight model Hy3 under permissive Apache license, outperforming GLM-5.2 on most benchmarks at roughly half the model size, though it trails in coding tasks. Why it matters: This...
Anthropic: Small firms quit Salesforce for Claude, signaling AI-native ERP shift
Small businesses are increasingly using Anthropic's Claude AI model to replace traditional enterprise applications like Salesforce, moving core business functions to AI-native workflows. The trend hig...
Amazon accelerates on-device AI chip design and plans screenless AI wearable
Amazon is accelerating its on-device AI chip strategy and signaling a major hardware pivot. During a CNBC podcast appearance, Panos Panay, SVP of Devices & Services, confirmed that Amazon now designs...
OpenAI VP argues for test-time compute as new evaluation standard, replacing single benchmark scores
Noam Brown, OpenAI's VP of Research, used a keynote at the Global AI Frontier Symposium 2026 in Seoul to argue that the AI industry must move beyond single benchmark scores for evaluating frontier mod...
Finatext Holdings has developed a practical model for responsible generative AI usage and has begun...
Finatext Holdings has developed a practical model for responsible generative AI usage and has begun implementing an automated assessment system. The Tokyo-based financial technology company announced...
Anthropic wants to develop its own drugs. At an event called “The Briefing: AI for Science,” the com...
Anthropic wants to develop its own drugs. At an event called “The Briefing: AI for Science,” the company announced Claude Science, an AI workbench for scientists that integrates fragmented tools and d...
Anthropic moves to close loopholes that let Chinese firms like Ant use its models via workarounds
Anthropic is implementing measures to close loopholes that allowed Chinese companies, including Ant Group, to access its AI models through workarounds involving cloud providers, according to Financial...
Alibaba Releases SkillWeaver Framework, Cutting Agent Token Consumption 99%
Alibaba has open-sourced SkillWeaver, a new AI framework that dynamically selects and loads only the relevant tools for a given agent task rather than loading the entire tool library. In complex multi...
Meituan's LongCat-2.0: First Trillion-Parameter Model Trained Entirely on Domestic Chinese Chips
Chinese food-delivery and local-services giant Meituan has released LongCat-2.0, a 1.6-trillion-parameter mixture-of-experts (MoE) model that claims to be the world's first trillion-parameter large la...
Anthropic cuts off Fable 5 and Mythos 5 access after White House export control directive
On June 12, 2026, the Trump administration ordered Anthropic to block foreign access to its latest AI models, Fable 5 and Mythos 5, citing national security export control authorities. Anthropic compl...
Apptronik opens Robot Park with Google funding for humanoid robot data collection
Apptronik, the Texas-based humanoid robotics startup backed by Google and Mercedes-Benz, today announced the opening of 'Robot Park,' a 90,000-square-foot facility in Austin, Texas, built to collect m...
Meituan, the Chinese food delivery and services giant, has released a new open-source large language...
Meituan, the Chinese food delivery and services giant, has released a new open-source large language model trained exclusively on Chinese-made chips. This marks the first known instance of a major Chi...
Meituan Releases LongCat-2.0, a 1.6-Trillion-Parameter Agentic Coding Model Trained Entirely on Chinese Chips
Chinese food-delivery and services giant Meituan has open-sourced LongCat-2.0, a 1.6-trillion-parameter agentic coding model released under the MIT license. The company states that the model, which ha...
Chitose Robotics tests VLM reference-information design for robot control program generation
Chitose Robotics (チトセロボティクス) has published results from a controlled experiment evaluating how different types of reference information fed to a Vision-Language Model (VLM) affect the quality of gener...
Elon Musk announces Grok 4.5 enters private testing at SpaceX and Tesla
Elon Musk has announced that xAI's next-generation model, Grok 4.5, has entered private testing at his other ventures, SpaceX and Tesla. The testing is being conducted internally to evaluate the model...
Wisorobotics releases ALLEX humanoid simulation model, starts building a Physical AI ecosystem
South Korean robotics firm Wisorobotics (위로보틱스) has released a simulation model of its ALLEX humanoid robot, allowing developers and researchers to study control, training, and synthetic data generati...