How AI agents are reshaping businesses & team dynamics | Microsoft CMO of AI at Work Jared Spataro

Summary

Jared Spataro argues that AI is becoming an on-demand business input—“intelligence on tap”—and that companies will increasingly organize work around teams of people and AI agents, which he calls digital employees. In his view, companies that learn to combine human judgment with AI execution will gain an advantage; those that fail to adapt risk disruption.

Business strategy and operating model

• Shift from individual execution to orchestration: Knowledge workers may increasingly act like managers—setting goals for agents, delegating work, evaluating outputs, and bringing results together. Spataro expects this to change roles across functions such as analysis, writing, and design.
• Rethink how firms allocate resources: Managers will need to decide where to use human talent, where to deploy AI, and how to coordinate both to produce business outcomes.
• Treat agents as digital labor: Spataro says this framing makes AI’s value easier to assess using familiar concepts such as employee capacity and fully burdened labor costs.
• Redesign teams and processes, not just individual tasks: He advises companies to move beyond isolated experiments and consider how AI could change each team and process.
• Build human oversight into deployment: Digital employees need to be onboarded, connected to relevant systems, given permissions and limits, and monitored as they ramp up. The interview does not portray agents as instantly autonomous across an entire company.
• Develop AI literacy across the organization: Spataro recommends building broad employee capability, then identifying specific workflows where AI can be applied.

Suggested implementation playbook

1. Increase AI literacy across the firm so employees and managers can identify viable uses.
2. Hire or deploy a first digital employee on a defined task or workflow.
3. Onboard and govern the agent: Connect it to appropriate systems, set boundaries, and monitor its performance.
4. Measure business impact using a clear baseline, such as avoided costs, added capacity, or revenue per employee.
5. Form human-agent teams and begin tracking the human-to-agent ratio as deployments expand.
6. Scale beyond pilots: Reassess teams and processes across the organization rather than keeping AI use localized.

Spataro’s advice to leaders is to begin immediately: “hire your first digital employee this week,” then use early experience to guide broader adoption.

Metrics, findings, and timelines cited

• 82% of surveyed people said they were confident they would use “digital labor” to expand workforce capacity in the next 12–18 months.
• 80% of the global workforce—employees and leaders—said they lacked enough time or energy to get their work done.
• Microsoft’s 2025 Work Trend Index drew on survey responses from 31,000 workers across 31 countries, LinkedIn labor-market and hiring trends, and Microsoft 365 productivity signals.
• Spataro said some Frontier Firms were generating 4×, 10×, or more revenue per employee than peers. These are examples he cited, not a general benchmark or guaranteed outcome.
• He proposed the human-agent ratio as a future measure of how organizations deploy AI, while noting that the appropriate ratio is not yet known.
• Spataro described an earlier Copilot productivity example in which reported time savings rose from 20 to 30 to 40 hours per month for effective users. He noted that CFOs found time savings difficult to translate directly into bottom-line impact.

Examples and case studies

• Dow: Spataro said the company expected a supply-chain agent to save millions of dollars in its first year by identifying misapplied fees.
• Supergood: The ad agency reportedly incorporated decades of advertising research into a platform intended to scale expertise across teams.
• AI-powered staffing firm: Spataro cited a one-person firm on track to earn $2 million that year, illustrating how AI-native businesses may operate with very lean staffing.
• Microsoft Work Trend Index team: Data scientist Alex Farach created three agents: one to find relevant research, one to assist with statistical analysis, and one to draft briefs connecting findings.
• Procter & Gamble field test: Spataro highlighted a study called The Cybernetic Teammate, co-authored by Harvard Business School professor Karim Lakhani. He described its finding as showing that one person equipped with AI could perform as well as a team without AI.
• Organizational design examples: Spataro cited Frontier Firms choosing not to hire a CFO or CMO, instead using agents to support analysis or giving an earlier-career employee AI tools to take on work that might otherwise require a more experienced hire.

ROI and business-model implications

Spataro described a progression in how companies assess AI value:

• Individual productivity: Early gains, such as time saved, were often hard for finance teams to connect to revenue or operating expenses.
• Process re-engineering: Applying AI to a specific process, such as customer support, can create more measurable cost impacts, but companies can only redesign a limited number of processes at a time.
• Digital labor: Comparing agent-supported capacity with employee costs may make value easier to quantify, because organizations already understand hiring, onboarding, and fully burdened labor costs.

He also suggested AI could change how some services are monetized. For example, if legal work can be completed faster with AI, billing primarily by the hour may become a less suitable business model.

Leadership and talent implications

• Managers may need new skills in workflow design, resource allocation, delegation, quality review, and coordinating human and AI work.
• Education and early-career development may need to emphasize how to direct agents, assess their work, and revise or combine outputs—skills Spataro says often take years to develop through workplace experience.
• The interview frames AI as a way to expand capacity and innovation as well as reduce costs, rather than focusing only on job replacement. It does not provide a quantified forecast of net job effects.

Sources and presenters

• Presenter/host: Molly Wood, WorkLab podcast.
• Guest: Jared Spataro, Microsoft Chief Marketing Officer for AI at Work.
• Research cited: Microsoft’s 2025 Work Trend Index; The Cybernetic Teammate study, co-authored by Karim Lakhani and conducted with a Procter & Gamble field test.
• Company results and predictions are reported as cited by Spataro in the interview.