Microsoft says that fiscal year FY26 saw customers across industries move from experimenting with artificial intelligence to using it to achieve tangible business outcomes, such as reducing time and costs, accelerating decision-making, and improving services. The company refers to organizations that place AI at the center of how they work as “Frontier Firms,” because they do not settle for efficiency gains, but use the technology to support human ambition and create new sources of value.
According to the review published on Microsoft’s official blog on July 28, 2026, the most successful organizations are building what the company calls an “intelligence platform” that brings together their knowledge, data, workflows, applications, and proprietary expertise. In this way, the value of AI accumulates within the organization itself rather than being concentrated solely in the model being used. Microsoft believes this approach also requires a trust platform capable of managing, governing, securing, and measuring AI across business processes.
Three Components of Frontier Transformation
Microsoft links its vision of frontier transformation to three main components: Microsoft 365 Copilot enables the use of AI within workflows and human ambition, while Microsoft IQ aims to strengthen and protect institutional knowledge, and Agent 365 provides a layer of trust and monitoring across the different levels of the ecosystem.
The company notes that AI systems do not remain static as organizations change. Agentic workflows therefore need to be continuously built, monitored, and tuned according to the outcomes the organization is targeting and the return on investment it requires. Microsoft says its open, model-diverse, and heterogeneous platform supports this improvement loop. It also recently announced Microsoft Frontier Company to help customers build AI systems that accelerate access to measurable business outcomes.
Examples of Agent Use in Organizations
Microsoft offers a range of use cases to illustrate this transition. Atos Group deployed Microsoft 365 Copilot to 56,000 employees in 54 countries and became one of the first organizations worldwide to adopt Microsoft 365 E7: The Frontier Suite. Using Microsoft Foundry, Microsoft Copilot Studio, and Agent 365, the company is building and operating a governed ecosystem comprising 19,000 AI agents within a unified model for productivity, security, compliance, and agent management.
In banking, Banco Popular Dominicano used the AURA ecosystem—a group of specialized agents built on Microsoft Copilot Studio and Microsoft Power Platform—to monitor the full scope of operational risk in real time, compared with approximately 40% previously. Microsoft says the system increased analytical capacity sevenfold, reduced manual operational effort by 70%, and achieved systematic accuracy of 98%, processing approximately 80,000 documents per week and more than 300 cases per day.
Chow Tai Fook deployed more than 400 dedicated agents to support over 24,000 employees, with millions of interactions per month, and achieved gains of more than 70% in the efficiency of core operations. The company says the AI Fook ecosystem helps frontline employees access product and inventory information and personalized recommendations, with an improvement in sales conversion reaching 57%.
In retail, Grandiose Supermarkets built the GrandChef shopping assistant using Microsoft Foundry and Azure OpenAI Service. The system connects meal inspiration, recipe discovery, and product purchasing using live product and inventory data. According to figures cited by Microsoft, the conversion rate rose 31%, average basket value increased 20%, and shopping journeys became 40% faster.
AI in Security, Healthcare, and Research
Other cases demonstrate AI’s role in sensitive environments. ASM used the Microsoft Security Copilot platform to accelerate threat investigations and protect intellectual property related to semiconductor manufacturing. The company says it reduced incident triage time by 68%, shortened investigations into laptop compromises from 25 minutes to eight minutes, and saved 337 hours per week, while reallocating 20% of security operations employees to governance, risk, and compliance initiatives.
At NHS England, a trial involving 30,000 workers across 90 organizations affiliated with the National Health Service resulted in an average saving of 43 minutes of administrative work per user per day. The authority subsequently began rolling out Microsoft 365 Copilot to more than 500,000 doctors and support staff, using Microsoft Copilot Studio to build agents for clinical, operational, and administrative workflows, and managing and scaling these agents through Agent 365.
Novo Nordisk, working with the Microsoft AI Acceleration Studio within the Forward Deployed Engineering team, also built a governed reasoning agent on Azure using a proprietary dataset containing more than 200,000 patient-years of curated clinical-trial data. The company says the system expanded its ability to evaluate strong ideas from 5 to 10 ideas per quarter to more than 50 ideas, and expects it to reduce the time required to reach insights in exploratory analytics from weeks to minutes.
From Productivity to Rebuilding How Work Is Done
Microsoft also cites EY’s experience, which deployed Microsoft 365 Copilot to 150,000 employees and achieved, according to the company, a 15% increase in productivity before expanding Microsoft 365 Frontier Suite to a global workforce of more than 400,000 people. EY also reported reducing completion times by 95%, cutting finance operating costs by more than 37%, and reducing manual work by up to 90% in key processes.
At Grupo Bimbo, the Audit Assist agent, developed using Microsoft Copilot Studio and deployed through Microsoft 365 Copilot and Microsoft Teams, helped reduce the audit-planning phase by 20% and accelerate the preparation of risk and control matrices from days to seconds. Navien, meanwhile, saved 28,000 hours annually through agents and automation, while 32% of its employees used self-service analytics to make decisions without IT support.
The examples also include Stellantis, which is developing more than 100 AI initiatives in sales, customer service, product development, and operations, and aims to reduce data-center space by 60% by 2029; and the University of Kentucky, which consolidated more than 150 AI initiatives under the CATS AI governance framework and made AI tools available to more than 70,000 students and employees.
Microsoft’s conclusion is that the next phase is not simply about adding an intelligent assistant to applications, but about building agentic systems that are continuously monitored and improved according to business outcomes. From the company’s perspective, institutional knowledge accumulates over time, while governance and trust enable usage to expand across regulated sectors such as healthcare, finance, and manufacturing. The performance indicators included in this review remain based on the cases presented by Microsoft and the experiences of its customers.