Public transcript

Microsoft CEO Nadella on How AI Can Change Workflows

By: Bloomberg Television
Length: 4 min
Language: English
Text from: the video's captions
Transcribed: 13 October 2026

Read the full transcript Open on YouTube

Notes in English.

Summary

Microsoft CEO Satya Nadella discusses how AI changes work, work artifacts and workflows, comparing it to the shift brought by the PC era. He describes how Copilot has inverted information flow inside Microsoft, flattening organizational structures. He argues that realizing AI's productivity gains requires a mindset change, new skills, and proper data context, and that results will vary by firm depending on leadership.

Key points

Nadella says the big challenge with new technologies is when work, work artifacts and workflows change, forcing firms to change how they work.
He recalls the CEO of Generali describing how the PC replaced faxes and interoffice memos and changed entire work processes.
Nadella now prepares for his roughly 50 bilateral meetings at Davos by asking Copilot for a brief instead of relying on his field team's notes workflow used since 1992.
Copilot gives him a 360 view, capturing information about clients, investments and relationships, which he shares back with colleagues across functions.
He says AI flattens information flow in organizations, inverting the classic model where information trickles up through departments.
His formula for taking advantage of AI is mindset, skills, and the data set or context you feed the intelligence layer, which some call context engineering.
He warns firms may not see immediate productivity results because they must do the hard work of organizational change, and differences will come down to leadership.
He stresses that leaders must use AI to learn how to put guardrails in place to trust it, rather than being afraid of it.

Questions it answers

00:00AI as a workflow revolution

Why does Nadella compare AI's impact to the arrival of the PC?

He says the big challenge of new technologies is that work, work artifacts and workflows change, forcing firms to change how they work. He recalls the CEO of Generali describing how the PC replaced faxes and interoffice memos and transformed their work with field agents.

New technology matters when it changes work, work artifacts and workflows.
The PC era replaced faxes and interoffice memos with spreadsheets sent by email.
AI is expected to drive a similar transformation.

00:57Preparing for Davos with Copilot

How has Nadella personally changed his meeting preparation workflow?

Instead of the field-team notes workflow used since he joined in 1992, he now asks Copilot for a brief before each of his roughly 50 bilateral meetings at Davos. It gives him a 360 view of the relationship, which he shares back with colleagues across all functions.

Copilot produces meeting briefs from event and relationship information.
He shares the brief with colleagues across all functions.
This inverts how information flows in the organization.

01:43Flattening the organization

What does Nadella say AI does to organizational structure?

He says AI flattens information flow, replacing the classic model where information trickles up through departments and specializations. Once that happens, the current structure may not make sense and firms have to redesign structurally.

Information flow is inverted and flattened by AI.
Current organizational structures may need to be redesigned.
Structures should allow information to flow freely.

02:37Mindset, skills and guardrails

What does Nadella say leaders need to do to adopt AI successfully?

He says the formula starts with a mindset of changing work and workflow with the technology, followed by skills. Leaders must use AI to learn how to put guardrails in place so they can trust it, rather than being afraid of it.

The formula starts with mindset, then skills.
You have to use AI to learn how to trust it.
Guardrails come from hands-on use, not fear.

03:27Context engineering and data

Why does Nadella say the data set matters for AI?

He explains the intelligence layer is only as good as the context given to it, which people describe as context engineering. Firms hold tacit knowledge across departments, and the challenge is giving AI that same context.

AI's intelligence layer is only as good as the context provided.
This is described as context engineering.
Firms' tacit knowledge must be fed to AI.

04:12Why productivity results vary

Why might firms not see immediate productivity gains from AI?

Nadella says firms have to do the hard work of changing mindset, skills and context, so results will not be immediate. There will be firmwide and sector-wide differences, but outcomes will fundamentally come down to leadership within the organization.

Immediate productivity gains are not guaranteed.
Hard organizational work is required.
Differences in results come down to leadership.

Notes

AI and workflow change

Satya Nadella frames the challenge of new technology as the moment when work, work artifacts and workflows change, which means firms have to change how they work.
He recalls meeting the CEO of Generali, who joined the firm pre-PC era and described how faxes and interoffice memos with field agents were replaced by spreadsheets sent by email, changing the entire work process.
He sees AI driving a similar kind of change in how work gets done.

A personal example with Copilot

Preparing for about 50 bilateral meetings at Davos used to follow a workflow unchanged since he joined Microsoft in 1992: his field team prepared notes, which were refined at HQ.
Now he asks Copilot for a brief before meeting someone, and it gives him a 360 view, capturing information about the relationship as client, partner and investment.
He shares that brief immediately with colleagues across all functions.

Flattened information flow

Nadella describes this as a complete inversion of how information flows in an organization, replacing the classic model where information trickles up through departments and specializations.
Once information flows freely, the current structure may no longer make sense and firms have to redesign structurally.

The formula: mindset, skills, data

Mindset: leaders should think about changing the work and workflow with the technology.
Skills: you have to use the tools to trust them and learn how to put guardrails in place, rather than being afraid of AI.
Data: the intelligence layer is only as good as the context given to it, which people describe as context engineering; firms hold tacit knowledge and must give AI that context.

Why productivity gains vary

Firms may not see immediate productivity results because they have to do the hard work of change.
There will be firmwide and sector-wide differences, but outcomes will fundamentally come down to leadership within the organization.

Text from the video's own captions. Summary and notes written by AI from the transcript, so check anything important against the recording.