AI agents are software systems that use a language model to plan a sequence of steps and then take actions — calling tools, writing files, querying databases — instead of only producing text.
Most business software is a series of small, repetitive decisions. If agents can handle even a slice of that reliably, the economics of back-office and support work change. The open question is reliability, not capability.
A model receives a goal and a set of tool definitions. It proposes an action, the system executes it, and the result is fed back into the context. A loop of plan, act and observe continues until the goal is met or a limit is hit.
Narrow agents work well: coding assistants, data extraction, customer support triage. Long-horizon autonomy remains unreliable, and most production deployments keep a human approving consequential steps.
Inference capacity, tool APIs, sandboxed execution and observability tooling.
AI safety company and developer of the Claude family of models.
San Francisco, United States · Founded 2021
Developer of the GPT models and ChatGPT.
San Francisco, United States · Founded 2015
Maker of the GPUs and software that dominate AI computing.
Santa Clara, United States · Founded 1993
Anthropic has released the newest version of its mid-range model, boasting faster response times and less token burn.
Source: TechCrunch · Read the original
Shopify is expanding WebMCP support to checkout, allowing browser-based AI agents to update order details and complete purchases with a buyer's authorization.
Source: TechCrunch · Read the original
Jensen Huang introduced software and hardware products that add independent security layers around AI agents so they stay within their environments even if they try to break out.
Source: TechCrunch · Read the original
Recent hacks have shown that the law is lagging when it comes to holding companies accountable.
Source: MIT Technology Review · Read the original
EvilTokens provided an end-to-end platform that makes mass account compromises faster and easier.
Source: Ars Technica · Read the original