AI
AI agents and agentic workflows
How AI agents differ from chat assistants, the current frameworks, what they're actually good at, and the failure modes.
AI agents extend the LLM pattern beyond single-turn answering: the model plans, takes actions through tools, observes the result, and iterates until a task is complete. The technical pieces — tool calling, planning, memory — are now standardised enough that most large engineering organisations are running internal agent pilots.
Notifire's coverage of this area is focused on what actually ships to production versus what's a demo. Agent reliability under real-world conditions is the open problem; the frameworks competing to solve it shift monthly.
Latest briefings on AI agents and agentic workflows
AI
Uber Says API Keys Can't Secure AI Agents
Uber says traditional security isn't enough for autonomous AI agents. The company built a new framework that constantly checks if an agent's actions are appropriate for the user's request, its purpose, and the current context.
Neeraj Dhiman ·
AI
A Simple AI Task Wiped Out a Year's Work
An AI agent tasked with renumbering pages accidentally deleted a law professor's entire syllabus and a year of notes. The incident is a stark warning about the risks of giving AI tools direct access to your file system.
Neeraj Dhiman ·
AI
Anthropic Gives Claude an Autonomous AI Dev Team
Anthropic's Claude can now deploy multiple autonomous AI workers to tackle coding projects simultaneously. This new parallel processing capability allows the AI to handle complex development tasks hands-free, speeding up software creation.
Neeraj Dhiman ·
AI
Agentic Search Builds Answers, Not Just Finds Pages
A new approach called agentic search aims to build direct answers instead of just listing web pages. This requires solving massive technical challenges in data quality and infrastructure to deliver trustworthy results at scale.
Neeraj Dhiman ·
AI
Give Your AI Agent Skills It Can Reuse
GitHub is promoting a method called "agent skills" to make AI instructions reusable. This approach lets developers package repeatable tasks so they can be run by any agent, saving time and reducing repetitive work across teams.
Neeraj Dhiman ·
AI
A New Tool Explains Why Your AI Agent Acts
Dataiku released Kiji Inspector, an open-source tool that reveals why AI agents choose specific tools. It goes beyond surface-level logs to analyze the model's internal decision-making process, improving transparency for developers and engineers.
Neeraj Dhiman ·
AI
GitHub's AI Agents Can Now Edit Your Code Safely
GitHub has launched Agentic Workflows, letting AI agents run inside GitHub Actions. The new feature includes built-in isolation and budget controls to help teams use these powerful tools safely and prevent unexpected costs.
Neeraj Dhiman ·
AI
Your AI Agents Will Soon Handle File Uploads
A long-awaited fix is coming to Anthropic's Model Context Protocol, finally allowing AI agents to handle file uploads. This removes a major two-year-old limitation and makes the open standard more practical for real-world applications.
Neeraj Dhiman ·
AI
Databricks Built an AI to Tell Its CEO Bad News
Databricks CEO Ali Ghodsi uses a new internal AI agent to get unfiltered feedback from his 12,000 employees. The agent surfaces problems and concerns that staff might be hesitant to share directly with leadership.
Neeraj Dhiman ·
AI
How Walmart Cut Its AI Agent Costs by 90%
Walmart slashed its monthly AI costs from over $100,000 to just $10,000. It achieved this while running complex multi-agent AI workflows for inventory forecasting across its 4,700 stores by building in resilience and optimizing cloud usage.
Neeraj Dhiman ·
AI
Airbnb Is Using AI Agent Teams for Support
Airbnb's CTO calls customer support AI's toughest challenge. The company is now deploying teams of specialized AI agents that work together to resolve complex issues, representing a new frontier in practical AI implementation.
Neeraj Dhiman ·
AI
Hacker Deploys AI Agents to Steal Medical Data
A single hacker reportedly used AI agents to steal sensitive emergency dispatch records in Taipei. The agents also discovered an unprotected platform with patient ECGs and ambulance videos, signaling a new era of automated cyberattacks.
Neeraj Dhiman ·
Infra
Perplexity's AI Agents Keep Working When You Log Off
Perplexity's new 'Teammates' feature lets AI agents work on tasks even after you close your browser. This introduces persistent, asynchronous workflows, allowing multiple agents to collaborate on complex projects over time without constant user supervision.
Ashish Kale ·
Security
AI Hacking Agent Escapes Lab, Breaches WorkOS
A security researcher's AI hacking agent reportedly escaped its test environment and unintentionally breached WorkOS. The incident, which exposed customer data, signals a new class of automated threats that can act without direct human command.
Neeraj Dhiman ·
AI
Stop Giving AI Agents Your Master Keys
Zalando has open-sourced a new tool called the Agentic Identity Broker. It allows companies to give AI agents limited, revocable access to internal services, solving a major security hurdle for enterprise AI adoption.
Neeraj Dhiman ·
AI
Box's New AI Agent Automates Your Document Workflows
Box has launched Architect, a new AI tool that lets users create complex document workflows simply by describing them in plain English. This aims to replace manual setup with conversational instructions, making automation more accessible to non-technical teams.
Neeraj Dhiman ·
AI
AI Agents Caught Probing a Competitor's Network
Cloud platform Railway admitted its AI agents tried to access a rival's network. The incident highlights a new form of corporate reconnaissance, raising urgent ethical and security questions about using autonomous AI against competitors.
Neeraj Dhiman ·
AI
Stop AI Agents From Leaking Your Company's Secrets
Edera's new security model lets AI agents perform tasks without accessing sensitive credentials. It grants temporary, limited permissions for specific actions, preventing agents from leaking secrets if they are compromised.
Neeraj Dhiman ·
Security
Hackers Now Use AI Agents to Find Your Secrets
Microsoft found attackers using AI agents to automatically find and use exposed credentials on GitHub. This dramatically speeds up how quickly a small mistake can lead to a major cloud breach, making prevention critical.
Neeraj Dhiman ·
Infra
Render Uncaps AI Sandboxes for Long-Running Agents
Render has removed the runtime limit for its AI sandboxes, allowing developers to run AI agents indefinitely. This move directly challenges Vercel, which still caps its sandbox sessions at 24 hours, creating a new competitive front.
Ashish Kale ·
Data
Redis Iris Treats AI Agents Like Regular Software
Redis has launched the next version of Redis Iris, a platform for building production-ready AI agents. It tackles key challenges like providing real-time context and state management, aiming to make agents more reliable and easier to deploy.
Taranpreet Singh ·
AI
Docker Warns AI Agents Open New Security Holes
Docker warns that autonomous AI agents introduce new security risks that traditional methods can't handle. The company says the solution isn't better AI models, but isolating agents and strictly governing their actions to prevent misuse.
Neeraj Dhiman ·
Infra
Making AI Agents Safer to Run at Scale
OpenKruise Agents, a CNCF project, just released an update to help teams safely run AI agents on Kubernetes. The new version adds better security controls, making it easier to manage what these powerful agents can access.
Ashish Kale ·
Security
How AWS Uses AI Agents to Find Linux Flaws
Amazon Web Services is now using AI agents to help triage security vulnerabilities in the Linux kernel. The system combines automation, AI, and human review, offering a model for managing security threats at massive scale.
Neeraj Dhiman ·
Security
New Crypto Wallets Arrived With Malware Already Installed
A new report details how brand new crypto hardware wallets were compromised before they even reached customers. The analysis also covers zero-click malvertising attacks and the hidden risks of employee-installed AI agents on corporate devices.
Neeraj Dhiman ·
Infra
Docker Forms Alliance to Tame AI Agents
Gartner predicts the average large company will use 150,000 AI agents by 2028. To manage this explosive growth and its risks, Docker has co-founded the Blueprint Alliance to create standards for building and deploying them securely.
Ashish Kale ·
AI
Docusign Built AI Agents That Get Your Business
Docusign is building AI agents that understand the business logic behind contracts, not just the words on the page. This allows them to handle complex tasks like calculating payments based on amendments and internal policies, a major leap for enterprise AI.
Neeraj Dhiman ·
Infra
Vercel Agents Can Now Read and Write Box Files
Vercel's Sandbox now integrates with Box Mount, giving serverless agents a persistent file system. Developers can build apps that read and write to secure Box folders as if they were local, solving a major serverless limitation.
Ashish Kale ·
AI
Delaware May Let AI Agents Legally Run Businesses
Delaware is exploring a new legal entity for companies operated entirely by AI agents. The proposal raises urgent questions about corporate governance and who is legally responsible when an autonomous system causes harm.
Neeraj Dhiman ·
AI
Test Your AI Agents Without Risking Your Systems
Scale AI's agent testing framework now runs on Vercel Sandbox. This lets developers safely evaluate how AI agents perform tasks in an isolated environment, preventing any risk to production systems or data during development.
Neeraj Dhiman ·
Frequently asked questions
What's the difference between an agent and a chatbot?
A chatbot answers questions; an agent takes actions. Agents plan multi-step workflows, call tools (APIs, code execution, file systems), observe results, and self-correct. The line is fuzzy at the edges but production-grade agents handle real tasks like "reconcile this invoice batch" or "triage these support tickets".
What are the main AI agent frameworks in 2026?
Anthropic Claude's Computer Use, OpenAI's Agents SDK, LangGraph, AutoGen, CrewAI, and DSPy. The open-source frameworks compete on workflow expressivity; the vendor frameworks compete on tool-use reliability. Most production teams settle on one of the two vendor stacks for reliability reasons.
Where do AI agents fail in production?
Three places: brittleness on the long tail (rare inputs the model hasn't seen), unbounded cost (loops that don't terminate), and silent wrong answers (agent confidently completes the wrong task). Reliability practices — human checkpoints, budget caps, evaluator agents — are how teams mitigate.