All articles tagged with #agentic ai
Nvidia CEO Jensen Huang says demand for compute power is growing exponentially as enterprises adopt agentic AI, calling Nvidia’s Blackwell platform and NVLink the current king of inference and signaling that the Rubin platform should deliver similar cost efficiency.
NIST launches the AI Agent Standards Initiative through its CAISI to develop industry-led standards and protocols that build public trust and enable an interoperable, secure ecosystem for autonomous AI agents. The effort includes an RFI on AI agent security due March 9 and a draft concept on software/AI agent identity and authorization due April 2, reflecting growing adoption of AI agents and ongoing concerns about safety and security across industries.
A MIT-led survey of 30 agentic AI systems finds pervasive lack of risk disclosure, minimal monitoring, and few stop options, signaling serious governance and security risks as agentic AI goes mainstream. The researchers urge developers to improve transparency, safety testing, and containment, noting limited company feedback and uneven safety practices across OpenAI, Perplexity, HubSpot, and others.
Meta and other tech firms are tightening controls on OpenClaw, the agentic AI tool, amid security and privacy fears; some ban its use on work devices while others cautiously test it on isolated machines and push for safeguards. OpenAI will keep OpenClaw open source, highlighting the tension between rapid AI experimentation and robust security governance across the industry.
Thomson Reuters launches CoCounsel Legal in the UK, delivering agentic AI that combines Deep Research from Westlaw and Practical Law with Tabular Analysis for large-scale document review, and integrates with Microsoft 365 and HighQ to embed AI across litigation, transactional work, and regulatory analysis.
Google AI and Peking University unveil PaperBanana, a five-agent agentic framework that turns raw methodology text into publication-ready diagrams and plots, combining a Linear Planning phase (Retriever, Planner, Stylist) with a 3-round Iterative Refinement (Visualizer, Critic); diagrams use image models and plots are produced as executable Python Matplotlib code to ensure data fidelity, and the system outperforms baselines on the NeurIPS 2025 PaperBananaBench in overall score, conciseness, readability, and aesthetics.
OpenClaw is an open‑source AI agent that can act within everyday apps, while Moltbook, a social platform for AI agents, has drawn millions of bot posts and sparked debates over consciousness and even religion. Researchers say the wave of agent interactions offers a rare laboratory for studying emergent behaviors, human–AI collaboration, and biases, though many observed “autonomous” actions are shaped by humans selecting models and personalities. Experts caution against anthropomorphizing agents or assuming true autonomy, noting that the dynamics mainly reflect human input and system design, with potential implications for future autonomous AI development.
A study authored by Vishal Sikka and Varin Sikka argues that large language models cannot perform certain computations or multi-step tasks beyond a certain complexity, effectively placing a hard ceiling on what ‘agentic’ AI can achieve and challenging claims of near-term autonomous AI or artificial general intelligence. While LLMs may improve, the research suggests they won’t exceed their computational limits or replicate true human-like intelligence in the foreseeable future.
Microsoft announced new agentic AI solutions to enhance retail operations, including intelligent automation across merchandising, marketing, and store management, with features like Copilot Checkout for seamless shopping, personalized shopping agents, catalog enrichment, and store assistant tools, aiming to improve efficiency, customer engagement, and decision-making in retail.
The article discusses the emergence of agentic AI and its potential impact on quantitative research, highlighting the transformative role of advanced AI technologies in the field, alongside promoting digital journalism subscriptions from the Financial Times.
Lovable, an AI software startup based in Stockholm, raised $330 million in a funding round that valued it at $6.6 billion, driven by surging demand for AI-assisted coding tools and 'vibe coding' applications, with major investors including Alphabet's CapitalG and Nvidia's venture arm, as the company expands its features and infrastructure.
Agentic AI is rapidly transforming the workforce by enabling professionals to work alongside autonomous AI agents, leading to new roles, organizational structures, and opportunities for innovation, with organizations needing to build robust data and management systems to thrive in this new era.
Designing successful agentic AI systems requires organizations to reorient workflows around outcomes, appoint mission owners, and foster cross-functional collaboration, moving beyond simple process tweaks to fully integrate AI into their operations.
Apple's former CEO John Sculley states that OpenAI is now Apple's main competitor in the AI space, highlighting that AI has not been a strong suit for Apple and emphasizing the need for the company to shift from the apps era to the agentic AI era, which involves autonomous, task-performing AI agents.
Perplexity has made its AI-powered Comet browser available for free to all users, allowing easy download and setup without a subscription. The browser features an agentic AI that can perform tasks on the internet on your behalf, such as opening links, summarizing pages, and more, offering a glimpse into the future of AI-assisted browsing.