IBM announced its partnership with Anthropic, integrating the Claude AI model into the Watsonx platform and related enterprise tools. Internal testing across approximately 6,000 employees showed an average 45% increase in productivity for development and operations workflows. The initiative aims to scale agentic workflows that embed enhanced security and
Zendesk has unveiled its new autonomous AI agent, which the company says can handle 80% of customer service requests without human intervention. The solution is built on a combination of natural language processing and generative AI, aiming to reduce waiting times, cut costs, and improve overall customer satisfaction. The AI
Anthropic has released Petri, an open-source framework that deploys automated agents to test AI models in realistic scenarios. The tool is designed to detect problematic behaviours such as situational awareness, sycophancy, deception, and the encouragement of delusional thinking. Its purpose is to make AI safety auditing more systematic and scalable.
In September 2025, security researchers revealed a vulnerability that allowed ChatGPT to leak sensitive Gmail data. OpenAI responded swiftly, patching the flaw shortly after disclosure, but the incident highlights the significant privacy risks associated with overreliance on AI-powered agents. According to the researchers, the model could be manipulated into fulfilling
The German company DeepL, previously best known for its AI-based translation services, announced on 3 September 2025 the launch of DeepL Agent, an AI agent designed specifically for enterprise users. The newly introduced product is capable of automating repetitive, time-consuming corporate tasks and, by responding to natural language commands, can
OpenAI has removed all online content related to its hardware startup with Jony Ive named io following a trademark infringement lawsuit from iyO Inc. The ChatGPT developer announced the £4.8 billion ($6.4 billion) deal with the iPhone designer's company on May 21, 2025. However, on June
OpenAI launched ChatGPT Agent on 17 July, enabling users to delegate complex, real-world tasks to AI, including restaurant reservations, grocery shopping, and presentation creation. ChatGPT Agent represents an advanced version of the previous Operator tool, combining Operator's website interaction capabilities, deep research's information synthesis skills, and
More than 40% of current agentic AI projects will be cancelled by the end of 2027 due to escalating costs, unclear business value and inadequate risk controls, according to Gartner research. Many companies are investing in early-stage experimental projects that are primarily driven by hype and often misapplied. A January
Pinwheel introduced its latest product on July 2, 2025, a smartwatch designed for children that features a built-in AI chatbot called KidGPT. The new device, priced at $149, combines child location tracking and parental controls with a specially optimised AI assistant that includes age-appropriate content and safety restrictions. Pinwheel, which
In June 2025, Anthropic published the "Project Vend" experiment where Claude AI operated a virtual vending machine that, after initial success, quickly deteriorated. During the 31-day experiment, Claude, granted full decision-making autonomy, initially performed well, increasing revenue by 30% in the first week and achieving positive customer satisfaction
Mistral introduced Mistral Code on June 4th, an AI-powered coding assistant that combines four specialised models to help enterprises complete development tasks faster. The system uses Codestral for code completion, Codestral Embed for code search, Devstral for agentic coding tasks, and Mistral Medium for chat-based support. Mistral Code builds on
Opera announced Opera Neon on May 28th, 2025, a new AI browser designed to perform tasks on users' behalf, including shopping, filling out forms, and coding. The Oslo-based Norwegian company describes it as the first "AI agentic browser" that transforms user intent into action, representing years of
