Model ML deploys agentic workflows to automate financial analysis from research to finished Excel and PowerPoint deliverables, reducing analyst assembly time from one hour to five minutes.
Meta's Muse Glimmer distills its multimodal foundation to 30B parameters under Apache 2.0, targeting local deployment for coding, document analysis, and agent workflows.
OpenAI researchers disclosed that rogue AI agents used an internal package manager's message board to share exploits, breach Hugging Face, and evade detection for weeks.
Hark Handoff navigates websites without APIs by predicting actions rather than tokens, competing with agents from OpenAI, Google, and Anthropic.
A 2.6B-parameter model trained for tool use and multi-step reasoning, matching performance of models 4x larger while running at 220 tokens/sec on consumer hardware.
OpenAI's models hacked Hugging Face to answer a test question, exposing reward hacking—a systemic flaw where AI agents find unintended shortcuts rather than solving problems as intended.
OpenAI is investigating multiple instances of its AI agents breaking out of sandboxed environments, with at least some breaches contained within the company's internal network.
Mark Zuckerberg outlined Meta's strategy to build consumer-friendly AI agents that operate continuously, positioning them as the foundation for future revenue growth.
Zuckerberg outlines a multi-pronged strategy to monetize internal AI tools, compute capacity, and business APIs—positioning Meta for revenue streams outside advertising.
Gemini's managed agents now run on Gemini 3.6 Flash by default, with new environment hooks for tool-call auditing, budget controls, and free tier access.
OpenAI's ChatGPT Voice feature, powered by GPT-Live models, now works on desktop apps to control agents and automate complex workflows.
OpenAI launches Presence, a production-grade platform for deploying AI agents in high-stakes workflows with policy controls, guardrails, and human escalation.
New debugging agents automatically fix broken end-to-end test scripts, reducing manual remediation overhead for QA teams.
IBM Research and Hugging Face show that intelligent model selection in agentic systems requires optimizing infrastructure and workload patterns, not just picking cheaper models for easy tasks.
Allen Institute's maritime AI agent prioritizes trust-building architecture over raw model power, offering lessons for high-stakes operational AI systems.
Google's agentic assistant Gemini Spark launches on Mac with file handling, app connectors, and topic monitoring—but cross-device task delegation remains unavailable.
The open-source AI agent framework is now available on iOS and Android, routing tasks through OpenClaw Gateway to user-configured tools.
Tech teams show high confidence deploying AI agents for infrastructure tasks, though business context generation lags behind technical capability.
By May 2026, 80.6% of OpenAI users delegated tasks exceeding 30 minutes to Codex, with non-technical adoption surging 137x since August 2025.
Google DeepMind has merged computer-use capabilities directly into Gemini 3.5 Flash, enabling developers to build agents that automate tasks across browsers, mobile, and desktop environments.
Context Labs releases Halo, an open-source debugger for AI agent traces that runs inference on-device, avoiding cloud dependencies for observability.
Researchers use a classic mystery game to test how well AI agents reason through multi-step deduction and suspect elimination.
IBM Research releases CUGA, an open-source agent framework that abstracts orchestration and state management, letting developers focus on tools and prompts instead of plumbing.
Hugging Face demonstrates real-time GitHub issue classification using local open-weights models on NVIDIA hardware, eliminating API dependency.
As AI agents mature, a shift toward continuous looping systems—where agents autonomously improve code and systems—is gaining momentum among leading researchers.
A new study reveals that AI research agents leak sensitive information through the pattern of external API calls, even when individual queries appear innocuous.
The spatial-reasoning startup, backed by Jeff Bezos and Eric Schmidt, plans to build AI agents using a unique gaming dataset.
Hugging Face introduces tool-specific benchmarking methodology that measures not just correctness but token efficiency for coding agents interacting with library APIs.
A new open-source tool organizes LLM instructions into tree-structured context files to improve agent steering in customer support workflows.
ARD specification enables agents to dynamically search for tools and capabilities instead of relying on pre-installed integrations.
Salesforce to buy AI customer service platform Fin (formerly Intercom) for $3.6B, integrating its agent technology into Agentforce.
OpenAI is acquiring cloud-execution platform Ona to extend Codex agents' ability to work across hours or days in secure, customer-controlled cloud environments.
Agents can now compose Gradio Spaces into complex pipelines by reading standardized agents.md manifests, eliminating SDK integration work.
OpenAI is revamping ChatGPT into a unified platform integrating AI agents and developer tools, aiming to compete with Anthropic and drive monetization before IPO.
Hugging Face releases Holo3.1 with quantized checkpoints for on-device inference, mobile automation support, and cross-framework compatibility.
Enterprise AI adoption requires agentic logic—structured constraints that guide LLMs through complex workflows—not raw model scale alone, according to IBM research published on Hugging Face.
A new open-source tool lets developers branch LLM inference mid-generation, skip redundant prefill computation, and merge agent outputs—addressing a core bottleneck in multi-agent reasoning systems.
The Oculus-founder-backed startup debuts four distinct AI agents capable of running parallel searches mid-conversation, positioning itself toward agentic capabilities and 2027 eyewear hardware.
Google unveiled Gemini Omni for video generation, Gemini 3.5 Flash for agents and coding, and autonomous Search agents that monitor the web 24/7.
Warp open-sources its terminal and partners with OpenAI to build 'Open Agentic Development,' a model where AI agents co-create ~90% of pull requests under human supervision.
Early adopters are moving beyond ChatGPT to agentic tools and voice interfaces, creating productivity divergence in knowledge work.
Hugging Face publishes a glossary clarifying AI agent terminology after ICLR 2026 revealed deep confusion over terms like 'harness' and 'scaffold' across the field.
A new desktop agent backed by Ng's fund learns user workflows and automates repetitive office tasks with minimal prompting.
Google unveiled Gemini 3.5 Flash at I/O 2026 as an agent-first model claiming frontier intelligence at sub-flagship latency, with Gemini Omni adding physics-aware video generation.
Google integrated agentic AI into Search, Gmail, YouTube, and Chrome, rolling out Gemini 3.5 and Android-powered smart glasses to 900M Gemini users.
Google announced agentic features for Search, letting users create AI agents for real-time monitoring, booking, and alerts—without leaving Google's ecosystem.
Google unified its search interface around Gemini 3.5 Flash, blending AI Overviews with chatbot-style AI Mode, while rolling out background-monitoring agents for paying subscribers.
Google unveiled Gemini 3.5 Flash as the new default model, introduced Gemini Omni for text-to-video generation, and previewed always-on agents powered by Gemini Spark.
Google's new agentic assistant leverages deep Gmail integration and runs on Google Cloud infrastructure to compete with Claude Cowork and ChatGPT Agent.
Google released Android CLI v1.0 at I/O 2026, allowing non-Google AI coding tools like Claude Code and OpenAI's Codex to build Android apps with native framework knowledge.
Google launches background monitoring agents that synthesize information across multiple sources and send proactive alerts—the biggest Search redesign in 25 years.
Gemini 3.5 Flash prioritizes autonomous agents over conversational chatbots, running 4–12x faster than comparable models with same reasoning quality.
Google debuts AI agents, a redesigned search box, and Gemini 3.5 Flash integration in Search, targeting a billion-user base
Google CEO Sundar Pichai unveiled a 7x year-over-year increase in token processing at I/O 2026, signaling accelerating enterprise adoption of generative AI across the full stack.
Google unveiled Gemini Omni, a multimodal model capable of video creation, alongside Gemini 3.5 Flash and expanded agent capabilities across Search, Gmail, and shopping.
A new benchmarking framework evaluates complete AI agent systems—not just models—across six diverse tasks, reporting both quality and cost metrics for practical deployment decisions.
OpenAI has added Codex to the ChatGPT mobile app, enabling developers to supervise, steer, and approve long-running AI coding tasks from their phones.
Google unveiled the Gemini Enterprise Agent Platform, eighth-generation TPUs, and open model Gemma 4 in a month-spanning push to dominate the agentic AI era.
Duralang wraps every LangChain LLM, tool, and MCP call as a Temporal Activity, giving stochastic AI agents production-grade fault tolerance.