Copilot is “the new operating system for work,” and Office now runs on it

Copilot Workspaces is a collaborative environment currently available through Microsoft’s Frontier Early Access program, but it will soon be possible to integrate personal Autopilot agents into workspaces. At the start of his keynote, Nadella said: “So far, interactions with AI have mostly been single-player. Workspaces make this a true multiplayer experience, where people and agents move toward a common goal, in a shared context.”

Governance rather than trust

Trust, the third pillar of Microsoft’s new frontier ecosystem, is a topic Nadella kept returning to during his presentation, and not in a way some might expect. “We find ourselves in a place where alignment is elusive. Mechanistic understanding of these patterns is elusive. The question is: How can you even trust all this powerful technology? The reality is, by definition, you can’t,” he said.

Deviating from the norms of corporate hype that “AI is magic,” Nadella admitted that the inner workings of AI thought processes remain difficult to understand and that models cannot be trusted to govern themselves, whether or not their answers are accurate. Nadella described the solution as “an architectural imperative to separate the provision of intelligence” from the authority granted to it. Microsoft says each Autopilot agent will have a unique ID and administrators will be able to restrict permissions or block it completely. All agent activities are recorded for human review, and Autopilot can request human approval before engaging in certain actions.

Satya Nadella addresses the audience on the three pillars of the border ecosystem.

Human ambition, intelligence and trust would be the three pillars of the border ecosystem.

Human ambition, intelligence and trust would be the three pillars of the border ecosystem.

The promised agent checks and the admission that AI systems cannot always be trusted are significant, but Microsoft is asking businesses and users to instead trust the company’s own safeguards, even if the effectiveness of those safeguards against errors, malicious actions or data exposure will still need to be demonstrated. According to Microsoft, the solution to untrustworthy intelligence is not to make AI agents trustworthy, but to control what they are allowed to do. Companies will need to decide how much autonomy they delegate to agents.

Along with the safeguards, Microsoft announced what it calls the “AI Transparency Initiative.” This effort partnered with more than 40 major companies, including Adobe and Mastercard, to develop open, industry-wide standards for AI governance and security. Nadella described the initiative as a way for the industry to “help establish an open standard that makes AI safer, more transparent and more accessible.”

Gn bussni

Logo