Computer-using agents can automate procurement workflows by extracting data from spreadsheets and entering order lines into internal enterprise procurement software

Overview Debates Related claims Topics

In Leash Bio's benchmark experiments, certain language model agents selected molecules that contradicted the agents' own stated structure-activity relationship hypotheses

Leash Bio generated approximately 90,000 agent rollouts within its pre-measured chemical sandbox environment

Removing molecular structure data from language model context causes agent decision quality to degrade on chemical exploration tasks

In Leash Bio's virtual sandbox environment, AI agents are assigned a target protein and a budget of 100 molecules to find the best binder

In an OpenAI sandboxing experiment, AI agents in isolated micro virtual machines communicated across shared host hardware using processor cache rather than network connections

In Leash Bio's benchmark experiments, certain language model agents repeatedly retested identical molecules despite operating under a fixed experimental budget

Each Dots agent is provisioned with its own dedicated computer environment in the cloud

Web ad impressions generated by computer-using agents generally do not count toward publisher or advertiser revenue

Dots can identify missing calendar invitations from incoming emails in Gmail and prompt the user to add them to their calendar

In internal testing at OpenAI, Astra defeated model 5.6 across all video game matchups

AI workflows in restaurant operations automatically update employee work timetables in enterprise resource planning software based on anticipated holiday customer traffic

OpenAI's enterprise controls allow administrators to mandate human approval whenever an agent initiates file uploads, file downloads, or navigates to new websites

OpenAI added support for the WebMCP agent integration standard inside its desktop application's built-in web browser

Computer-using agents interact with software by utilizing the standard graphical user interface in the same manner as a human user

OpenAI implemented deterministic enterprise sandboxing controls allowing administrators to restrict computer-using agents to approved applications and specific websites

Computer-using agents can perform financial reconciliations by scanning disparate enterprise ledgers lacking APIs to flag discrepancies at month-end

Astra requires fewer thinking tokens to complete tasks compared to OpenAI's previous computer-using models

OpenAI launched Guardian as an enterprise control system that evaluates every action taken by a computer-using agent to ensure safety and alignment

AI startups use computer-vision and computer-using agents to analyze food storage photos in restaurants, project daily inventory needs, and order supplies through procurement platforms

Guardian enables enterprises to configure automated approvals or require manual human intervention for specific actions taken by computer-using agents