As Scout VP Omar Shahine put it, the idea is to create an assistant that actively adapts to the user’s needs. “We all have our interesting quirks in how we work, and people are codifying those patterns into memories and skills that persist in their agent,” Shahine told me. “Then the agent becomes more capable, better understanding you and gaining more agency and exercising judgments.”

Available through Microsoft’s Frontier program, which gives early adopters access to experimental Microsoft products, Scout will require a GitHub Copilot subscription to use.

Scout is based in the cloud, but operates across the desktop and web browser also, so it’s easy to connect to inboxes, calendars and other systems. Scout will come with prepackaged skills for calendar management and drafting meeting agendas, among others, but Shahine expects the real value to be in the skills users develop on their own. That customization loop — where the assistant learns from user behavior and becomes more capable over time — is the same dynamic that has made consumer AI tools sticky; the more you invest in training your assistant, the harder it is to walk away.

The system also comes with extensive security protections, meant to address concerns of unsupervised AI agents running amok, a real issue that OpenClaw itself surfaced earlier this year when one agent was reported to have acted erratically inside a researcher’s inbox (among other examples). Scout will come with a built-in “policy conformance system” that will continuously check whether the system is operating according to set guidelines, and each conformance check will produce its own audit trail.