How can companies use new AI capabilities in their day-to-day work while maintaining control over data, costs and security? This question shaped the main direction of Microsoft’s announcements this year. At Microsoft Ignite 2026, the company showcased developments in Copilot, the Microsoft IQ platform, new models, agent management tools and security solutions. For organizations, the key issue is not simply the arrival of new features, but how they can be connected to existing processes and deliver measurable value.
Copilot: from answering questions to collaborating on tasks
Microsoft is developing Copilot as a shared work environment where users can get answers, delegate tasks or work on them together with an agent. The solutions presented include Copilot Workspaces for collaboration, the Autopilot agent for longer and repetitive tasks, and Copilot Code, which lets users create simple applications, dashboards and tools using natural language.
For a company, this could mean less manual transfer of information between applications and an easier way to introduce automation into teams’ day-to-day work. It makes sense to start with specific processes: identify which tasks can be delegated, what data is needed and who approves the agent’s actions.
Microsoft IQ: company knowledge as useful context
An agent may be capable, but without access to the right context, it will not understand the specifics of a company. Microsoft IQ is designed to bring together information from Microsoft 365, Fabric, Foundry and selected external systems, then make it available to agents in a controlled way. Business plugins being developed in this area are intended to let teams package knowledge, skills and tools into reusable components.
In practice, data quality and organization matter, as do clear access rules. The better a company defines its processes, roles and information sources, the more likely an agent is to support employees with a relevant answer or action. This is also a good opportunity to review where organizational knowledge is currently scattered.
Models: the right tool for the right task
The announcements include models with different purposes and cost profiles. Microsoft Decision 1.1 is intended to support tasks such as classification, routing cases to the appropriate workflows and managing workflow logic. Microsoft also introduced Code 2 and Code 2 Flash for programming tasks, as well as Cyber 2 Flash for cybersecurity use cases. Microsoft Foundry is designed to provide access to multiple models, allowing organizations to choose those that best fit their needs.
For organizations, this means considering AI not as one universal solution, but as a set of tools suited to different tasks. When choosing a model, it is worth comparing output quality, speed, cost, data requirements and how it can be integrated into existing processes.
Agent 365: visibility, accountability and costs
As more agents are introduced in a company, practical questions arise: who is responsible for them, what data and tools can they access, what risks do they pose, and how much do they cost? Agent 365 is intended to extend existing management, identity and security controls to agents. The announced Business Analytics features are designed to show usage and spending, and help connect them to outcomes.
This is an important step towards more structured AI adoption. The number of agents in use does not, by itself, show whether an investment is worthwhile. Organizations need metrics that can show, for example, time saved on a task, how successfully it was completed, the cost of handling it and the impact on work quality.
Security: agents supporting security teams
Microsoft Perception is designed to coordinate the work of specialized security agents. “Red” agents help simulate and identify attack paths, “blue” agents support detection and analysis, and “green” agents help remediate issues and reduce risk. The solution is intended to bring together signals from different parts of the IT environment, while security teams retain the ability to oversee its actions.
In security, response speed matters, but so does being able to check why a system took a particular action and who approved it. Automation can relieve specialists of repetitive tasks, but escalation rules and responsibilities should be defined before deployment.
AI transparency: making system behaviour verifiable
Microsoft announced the AI Transparency Initiative, a collaboration with organizations and partners to develop shared standards for transparency, security and the governance of models and agents. The initiative includes developing industry standards, working across the wider ecosystem and incorporating the resulting practices into Microsoft products.
For business users, the principle is straightforward: an AI deployment should make it possible to check what information a system uses, what actions it takes and how those actions can be controlled. The more an agent influences operational processes, the more important it becomes to log its activity, set appropriate permissions, test it and make accountability clear.
The future of AI in Business Central
The Ignite announcements point to a future in which AI is more deeply embedded in the processes, data and tools employees use. To move from demos to business value, organizations should start with a clearly defined problem, well-organized data, access controls and ways to measure results. Only then should they choose the model, agent and integration approach.
It will also be important to check which features are available, which are still in preview, and what licensing and technical requirements apply in each environment. Feature availability and scope may depend on region, plan and service configuration.
Want to introduce AI in your company?
At IT Vision, we help companies develop their Microsoft Dynamics 365 Business Central environment and connect it with the processes and data they use across their organization. If you are considering where to start with AI, let’s talk about the processes worth improving and how to move safely from an idea to a practical implementation.
AI for business with Microsoft: questions and answers
What is Microsoft IQ?
It is an organizational context management platform that brings together a company’s data, knowledge and processes so that AI agents can work with its information and rules.
How does Work IQ use ERP and CRM data?
Business Applications in Work IQ gives agents access to supported ERP and CRM data, combining it with work context such as emails, documents and meetings. The data available depends on the configuration and permissions.
Can Microsoft IQ plugins work in Claude?
Yes. Microsoft has announced that business plugins may also be used with external assistants, including Claude, subject to defined access and governance controls.
How can Agent 365 help assess the results of an AI deployment?
Agent 365 Business Analytics is intended to show AI usage and costs, and help connect them to business outcomes so organizations can assess the value of their deployment.
Where should companies start when deploying AI agents?
Microsoft recommends choosing one specific process where a better decision or action could deliver a measurable result, then testing the solution in practice.

IT Vision is an experienced provider of ERP systems, BI solutions, and B2B platforms, operating on the market since 2000. The company has completed over 400 projects worldwide, supporting organizations in their digital transformation. IT Vision’s team of experts combines business and technological knowledge, delivering high-quality implementations based on Microsoft technologies.



