Microsoft Intros’Cowork’Feature for Copilot, AI Updates
Microsoft has introduced a trio of AI updates covering Microsoft 365 Copilot, Security Copilot, and Microsoft Foundry.
The top news was the announcement of Copilot Cowork, a new mode in Microsoft 365 Copilot that Microsoft says is designed to assist users hand off multistep jobs instead of just getting answers back in chat.
“Copilot Cowork is developed for that: It assists Copilot take action, not just chat,” stated Charles Lamanna, president of Business Applications and Representatives.
Copilot Cowork Moves Beyond Chat
Cowork is designed to take a user’s objective and turn it into a structured plan that runs in the background. Microsoft said it pulls context from across Microsoft 365(Outlook, Teams, Excel, files, and meetings) through what it calls Work IQ, then surfaces checkpoints for approval before making modifications.
According to Microsoft, Cowork can deal with tasks such as solving calendar disputes, preparing meeting briefs, compiling research study memos with citations from web and office sources, and more.
Microsoft said Cowork operates within the existing Microsoft 365 security and governance structure, with identity, approvals and compliance policies imposed by default. Actions and outputs are auditable, the business stated.
Microsoft likewise said Cowork can tap Claude from Anthropic, which Lamanna referred to as a “multi-model benefit” that allows Copilot to route work to the design best matched for the task.
Cowork is presently offered to a limited set of customers through a Research study Preview. Microsoft stated it anticipates more comprehensive rollout through its Frontier program in late March 2026. The business introduced Frontier previously this year as an early-access channel for emerging Copilot features.
Security Copilot Gets a Credential-Finding Upgrade
Microsoft also announced the basic accessibility of Agentic Secret Finder, or ASF, in Microsoft Security Copilot.
The feature is aimed at identifying exposed qualifications hidden in unstructured data such as emails, chat logs, files and screenshots. Microsoft said ASF uses a multi-step, multi-agent reasoning procedure to figure out whether a suspicious string is a legitimate credential and what level of gain access to that credential might supply.
“Unlike regex-based scanners, ASF uses reasoning to identify not simply qualifications, but the systems they open, helping security teams comprehend exposure and react much faster,” Microsoft composed in the announcement.
Microsoft said that approach is planned to improve triage by decreasing the false positives often produced by conventional pattern-matching tools, while likewise recognizing credentials that do not fit known formats.
In benchmark testing utilizing synthetic datasets throughout e-mails, talks, notes, and files, Microsoft stated ASF reached 98.33% credential recall with no incorrect positives. Standard regex-based tools, the company stated, identified about 40% of the same credentials.
ASF presently supports more than 20 credential types, consisting of Azure Storage Keys, AWS Gain Access To Keys, OAuth tokens, SSH personal secrets, and database connection strings. Microsoft stated it is likewise checking out GitHub combination to extend the ability into source code analysis.
Fireworks AI shows up in Foundry
The company’s 3rd announcement was a public sneak peek that brings Fireworks AI to the Microsoft Foundry model catalog.
Microsoft stated the integration provides developers access to Fireworks AI’s cloud-based inference engine inside their Foundry jobs, using low-latency reasoning for a number of open source models.
“For customers needing the most recent open source designs from emerging frontier labs, break-neck speed, or the capability to deploy their own post-trained custom models, Fireworks provides best-in-class inference efficiency,” Microsoft said in the announcement.
At launch, the preview supports both serverless pay-per-token deployments and provisioned throughput across 4 designs: Minimax M2.5, OpenAI’s gpt-oss-120b, MoonshotAI’s Kimi-K2.5, and DeepSeek-v3.2.
Microsoft stated clients can also import and release their own fine-tuned versions from those model households– consisting of Qwen3-14B and DeepSeek v3.1– through a brand-new Custom Designs workflow in Foundry.
The Fireworks integration is opt-in throughout sneak peek and must be made it possible for through the Azure portal’s Preview features panel. Microsoft said consumers also must be in among 6 supported U.S. regions to use the pay-per-token option.