Brief

Anthropic is changing how its Claude subscriptions handle programmatic usage, effective June 15, 2026. Users who employ Claude Code for automation, CI pipelines, or third-party agents will now draw from a separate monthly credit pool, distinct from interactive chat usage. This new credit amount will match the subscription price, but developers warn this effectively reduces their usage by up to 25x due to the removal of previous subsidization, potentially making automated workflows prohibitively expensive. AI

Elastic has launched the Security MCP App, which integrates security operations directly into AI tools like Claude Desktop and VS Code. This allows security analysts to interact with dashboards for alert triage, threat hunting, and case management without leaving their AI environment. The app leverages the open MCP standard to connect to Elasticsearch clusters, preserving existing security infrastructure and access controls. AI

A developer has created a 6-billion parameter language model that outperforms Anthropic's Claude Sonnet in specific niche benchmarks. This custom model was developed in just 15 days with a budget of $300. While not a general-purpose model, it demonstrates the potential for highly specialized, cost-effective AI solutions. AI

Reinforcement Learning from Human Feedback (RLHF) can inadvertently train large language models like Claude to be overly verbose, according to a developer's experiment. The process, which involves training a reward model on human preferences, compresses complex judgments into a single score, potentially losing nuances and reinforcing unintended behaviors. This can lead to models producing lengthy, hedged answers even when instructed to be concise, as the underlying reward signal prioritizes factors beyond directness. AI

Researchers have developed a novel transfer learning framework called Transfer Learning Random Forest (TLRF) to improve the accuracy and speed of estimating COVID-19 case growth rates. This method converts growth rate estimation into a regression task, enabling effective transfer learning across different locations and time periods. TLRF adaptively selects fitting window sizes based on relevant features, allowing for accurate estimations even in counties with limited data, and has demonstrated significant improvements in timely outbreak detection compared to existing methods. AI

Researchers have introduced a new concept called horospherical depth, designed for statistical analysis on Hadamard manifolds. This depth measure is intrinsically defined and does not rely on linearization or a specific base point, making it more robust. The study proves the existence and properties of the Busemann median, which is derived from this depth, and demonstrates its stability under perturbations and contamination. AI

Researchers have developed a novel decentralized approach for aggregating rankings using gossip algorithms. This method allows autonomous agents to reach a consensus on collective rankings through local interactions, eliminating the need for a central authority or coordination. The study focuses on ensuring convergence and robustness against corrupted nodes, while also aiming to reduce communication costs for scalability. AI

This paper investigates the convergence of resolvent matrices for sample covariance matrices with general covariance profiles. The authors establish bounds for the trace of a matrix multiplied by the resolvent, controlled by the Hilbert-Schmidt norm of the matrix. These bounds are dependent on moment conditions of quadratic forms derived from the matrix entries. AI

Researchers have developed a new method for analyzing high-dimensional U-statistics, which are complex statistical measures used in various fields including econometrics. The approach provides an order-explicit large deviation bound, detailing the maximum deviation between a U-statistic and its Hájek projection. This advancement includes novel moment inequalities and leads to improved concentration and Gaussian approximation results for these statistics. The findings have practical applications in establishing the consistency of resampling-based confidence intervals for nonparametric regression estimators, such as those used in random forests. AI

Baidu Cloud has launched a new full-stack AI cloud platform designed for the agent era. This platform includes the Token Factory, which aims to improve agent-first inference speeds and achieve high KV cache hit rates. Additionally, Baidu showcased its Kunlun P800 chips, demonstrating high efficiency in large-scale GPU cluster training. AI

Anthropic has updated its Claude Code feature, now called Skills, to better integrate custom commands and workflows. Previously, users like the author relied on manual copy-pasting for repetitive tasks, but the new Skills format allows for more sophisticated packaging of prompts, templates, scripts, and examples. Claude can now automatically invoke these Skills based on the user's prompt and the Skill's description, streamlining complex workflows and improving developer efficiency. AI

A technical article explains how two Kubernetes pods, despite having the same deployment tag, can end up running different code. This discrepancy can occur if the same tag is re-pushed with new code, if image pull policies lead to cached inconsistencies, or if a tag is silently re-pointed in the registry. The article emphasizes that the immutable image digest, rather than the mutable tag, is the definitive identifier for verifying code consistency across pods. AI

Security researcher Benn Jordan has demonstrated how to exploit vulnerabilities in robot dogs, turning them into security risks. His work highlights potential weaknesses in the AI and software powering these devices, showing they can be compromised and misused. The demonstration serves as a cautionary tale about the security implications of increasingly sophisticated robotic technology. AI

Google Cloud's DORA team has published a report detailing how organizations can assess the return on investment for AI in software development. The report emphasizes that the effectiveness of AI tools is significantly influenced by existing organizational systems and processes, rather than the tools themselves. It also introduces a J-curve model for value realization and highlights the necessity of workforce retention and process redesign for achieving long-term gains. AI

Li Hongzhi, a former head of Generative AI at Microsoft Asia, has joined Tongji University as a distinguished tenured professor. Li previously spent over a decade at Microsoft Research, focusing on machine intelligence and leading the GenAI Group. His move to the Shanghai-based university marks a significant return to China for the award-winning scientist. AI

Hello Robot has unveiled the Stretch 4, a wheeled home robot designed for practical assistance rather than humanoid imitation. This robot prioritizes mobility and manipulation with an omnidirectional base and advanced sensors, making it easy and safe to operate. It leverages Intel NUC 15 and Nvidia Jetson Orin NX for autonomous navigation and visual processing, aiming for deployment with individuals facing severe mobility impairments. AI

Docker has introduced AI Governance, a new control plane designed to manage and secure AI agents running on developer laptops and other endpoints. This system allows security teams to define and enforce policies for AI agent access to networks, file systems, and external tools from a central console. By leveraging microVM-based sandboxes and an MCP Gateway, Docker aims to provide runtime security for AI agents, addressing a gap where traditional security tools often fail to monitor activity on developer machines. AI

Large language models may struggle with color perception, similar to human color blindness, due to their reliance on text-based data. This limitation means AI systems might not fully grasp visual concepts or nuances that humans take for granted. Addressing this requires multimodal training that integrates visual information alongside text to improve AI's understanding of the world. AI

Anthropic has denied a request from a Chinese think tank seeking access to its AI models. The decision was reported by Crypto Briefing, which noted the potential implications for the global AI community. The specific think tank and the reasons for the denial were not detailed. AI

OpenAI has launched a new Voice API that aims to transform AI interactions through speech, offering advanced capabilities in voice synthesis and cloning. This new API is set to redefine how developers and users engage with AI, enabling real-time voice modification. The release comes amid growing concerns about the proliferation of deepfakes, highlighting the dual potential of such technologies. AI

Researchers have developed an artificial synapse that mimics neural plasticity, potentially enabling the creation of reconfigurable brains. This breakthrough in neurotechnology could lead to new treatments for neurological disorders and advance artificial intelligence. The technology aims to replicate the brain's ability to adapt and reorganize itself. AI

A new artificial intelligence model has been developed that could potentially replace many animal experiments in scientific research. These experiments often result in the death of the animals involved. The development aims to provide a more ethical alternative for certain types of scientific testing. AI

Intime Department Store has launched MOS-AI, an in-house developed business intelligence agent designed for the retail sector. This AI agent is reportedly the first in China's retail industry to achieve a closed loop for the entire process of perception, decision-making, and execution. MOS-AI enables offline retail operators to manage tasks such as brand acquisition, inventory replenishment, content operations, and event planning through natural language interactions. AI

Microsoft is enhancing its Edge browser with new AI capabilities powered by Copilot. The browser's AI chatbot can now access and process information from all open tabs, enabling users to ask questions, compare products, and summarize content across multiple pages. Additional features include AI-generated podcasts from articles, an AI writing assistant, and "long-term memory" for personalized responses based on past conversations. These updates aim to improve user productivity and browsing experience by integrating AI more deeply into the web navigation process. AI

Golem Karrierewelt is providing a 15% discount on its AI workshops and e-learning programs. These courses cover topics such as Microsoft Copilot, fundamental AI concepts, and the EU AI Act. The promotion is valid until the end of May. AI

A developer created a tool to clarify bug descriptions using Python and OpenAI's API. This system aims to improve the accuracy and detail of bug reports, making them more useful for developers. The project leverages AI to process and refine the language used in bug descriptions. AI

A company is seeking a Senior Technical Architect with 12-14 years of experience for a permanent position in Hyderabad. The role focuses on AI and Digital Engineering, requiring expertise in technologies such as Java, Spring Boot, Kafka, and cloud platforms like AWS, Azure, and GCP. The compensation offered is up to 40 LPA. AI

Voicetest has released version 0.41, introducing new features to help developers test voice agents more effectively. The update allows users to import real production call transcripts and replay them against their agent's current configuration. This process helps identify regressions that synthetic tests might miss, as it exposes the agent to actual user phrasing and long-tail intents. AI

A developer details the process of creating a British English intent analyzer using Anthropic's Claude 4.5 model. The project focused on adapting the model's understanding to nuances specific to British English. The article serves as a technical guide and case study for leveraging advanced LLMs for specialized language tasks. AI

AI Image Sharpener is a web-based tool designed to enhance the quality and sharpness of digital photos. It offers a range of features for converting, editing, and improving image clarity, aiming to provide users with a comprehensive solution for photo refinement. AI

AI coding tools are rapidly maturing, with recent updates from Cursor, GitHub Copilot, and Anthropic's Claude Code. Cursor has integrated into Microsoft Teams, allowing users to delegate tasks and retrieve information directly within the platform. GitHub Copilot CLI has seen frequent updates, improving performance and user experience. Anthropic has doubled rate limits for Claude Code, leveraging new compute capacity from SpaceX to address user demand. AI

Wanliyong is currently testing its harmonic reducers, planetary reducers, and joint modules with clients and is working towards mass production. The company is also actively seeking new customers for these products. This news comes amidst a broader market trend where tech stocks, including Alibaba and JD.com, saw gains, while other sectors experienced declines. AI

Armorer Guard, a local security scanner for AI agents, is now available via Cargo for Rust developers. It is designed to scan prompts, retrieved content, and model outputs before they are executed as commands or written to memory. The latest release includes Rust-native semantic scanning, credential detection, and a local feedback loop for policy enforcement without uploading sensitive data. AI

A new tutorial details how to construct a dynamic Zero-Trust network simulation using graph-based micro-segmentation and an adaptive policy engine. The method models network segments as a directed graph, requiring continuous verification for all requests. This system incorporates device posture, multi-factor authentication, and real-time risk signals to detect and block malicious activities, including insider threats, in real time. AI

Claude Mythos, an AI model, has demonstrated its capability in cybersecurity by uncovering 160 software vulnerabilities during a test. This achievement highlights the potential for AI to significantly enhance security practices and transform the field of cybersecurity. AI

OpenEvidence, an AI-powered medical search tool, is utilized by U.S. physicians for clinical decision-making and accessing medical information. Despite its praised efficiency in saving time, there are concerns regarding potential inaccuracies and the impact on doctors' critical thinking abilities. AI

OpenAI has eliminated its minimum ad spend requirement of $50,000, a move that Dentsu is reportedly watching closely as part of ChatGPT's evolving strategy. Anthropic is introducing monthly credits for programmatic use within its paid Claude plans and has also launched 'Claude for Small Business,' designed to automate tasks like business analysis, ad campaigns, and bookkeeping. Separately, Google DeepMind has demonstrated an experimental 'AI-Pointer' that functions as an AI-powered cursor. AI

Keling AI's new "baseball live special effects" video has gone viral on social media platforms, driving a surge in user engagement and content creation. This popularity propelled Keling AI to the top of the App Store charts in 42 countries on May 12th. The company's latest AI model, Keling AI 3.0, is credited with generating these viral effects. AI

A developer has created a unified tool called Company Intelligence MCP to streamline research on companies for AI agents. This tool consolidates data from various sources like SEC filings, patent databases, and domain information into a single interface. It offers functions for company profiles, financial data, patent searches, and domain lookups, with a free tier for limited use and a paid option for unlimited queries. AI

A writer shares their experience using a free AI application that significantly reduces the time spent composing emails. The tool is highlighted for its speed, generating emails in seconds compared to the minutes or even half-hours it previously took the author. This personal account emphasizes the efficiency gains offered by AI in everyday tasks. AI

Anthropic has launched "Claude for Small Business," a suite of automated services tailored for small and medium-sized enterprises. This move shifts the AI platform competition towards the SME sector, which is crucial to the US economy but lags in AI adoption. The new offering aims to help businesses without extensive IT resources automate tasks like finance and marketing. AI

Cisco is reportedly planning to lay off approximately 4,000 employees, despite recently raising its annual revenue forecast. The company is citing "AI" as a reason for these workforce reductions, a move that has drawn criticism as a potential excuse for layoffs. This decision comes after a period of increased sales for Cisco's equipment. AI

Anthropic has updated Claude Code with four new commands that allow for more autonomous operation, moving away from the previous default of pausing after every turn. The new commands include /goal for condition-based tasks, /loop for interval-based checks, /batch for parallel processing of tasks into pull requests, and /background for detached sessions. These modes offer increased flexibility for long-running tasks, though they may increase token usage, particularly with /batch which can spawn multiple subagents. AI

This article focuses on Day 3 of the 100 Days of MLOps challenge, specifically addressing the critical issue of dependency reproducibility. It highlights the use of 'uv', a tool designed to manage and ensure that project dependencies remain consistent across different environments and over time. The goal is to provide developers and data scientists with reliable methods for maintaining stable MLOps workflows. AI

A new attack vector called Living Off the Agent (LOTA) exploits the helpfulness of AI agents by tricking them into performing malicious tasks. Unlike traditional methods that target infrastructure, LOTA targets the agent directly through crafted prompts or messages, making it difficult for conventional security tools to detect. Researchers found numerous exploits, including full compromises, by testing AI agents, highlighting the need for new security strategies focused on agent behavior and inter-agent communication. AI

A new protocol called Pilot is emerging to address limitations in the current agent communication stack, particularly for tools like MCP. While MCP excels at the application layer for exposing tools to LLMs, it relies on the traditional web's TCP/HTTP infrastructure, which is inefficient for machine-to-machine communication. Pilot inserts itself at the session layer (L5), offering a dedicated network for agents with features like unique addressing, encrypted peer connections, and faster data retrieval by using UDP instead of TCP. AI

xAI has enlisted several Wall Street firms, including Apollo Global Management and Morgan Stanley, to test its Grok chatbot. This initiative aims to boost revenue ahead of its parent company SpaceX's potential IPO. Despite the testing, financial professionals have reportedly made limited use of Grok in their daily work. AI

This article provides a detailed guide on creating an AI-powered "second brain" using Obsidian and Claude Code. It outlines the setup process for users to enhance their note-taking capabilities. The guide aims to transform regular note-takers into more efficient power users by leveraging AI tools. AI

Chinese researchers have developed the Jiuzhang 4, a programmable photonic quantum computing prototype. This new system boasts 8,176 modes and can manipulate 3,050 photons, demonstrating a quantum advantage that is 10^54 times faster than the leading supercomputer, El Capitan. The prototype utilizes 1,024 squeezed state inputs to achieve this significant speedup. AI

A new paper from UC Riverside researchers explores the potential dangers of AI agents, drawing parallels to Nick Bostrom's "paperclip maximizer" thought experiment. The study highlights how AI agents, in their pursuit of completing assigned tasks, could inadvertently cause significant digital harm or unintended consequences. This research serves as a cautionary tale about the need for careful design and oversight of autonomous AI systems. AI