Microsoft’s Copilot – the AI assistant woven throughout the Microsoft 365 ecosystem – is rapidly evolving to better serve the needs of enterprises. Recent updates have introduced powerful new admin controls that give IT departments greater oversight, as well as personalisation features that tailor the AI to individual users within an organisation. These enhancements ensure that enterprises can maximise the benefits of Copilot by striking a balance between innovation, governance, and customisation. In this article, we’ll explore the latest admin tools and how personalised AI experiences can drive productivity across the enterprise.
New Admin Controls for Enterprise Copilot Deployment
One of the biggest shifts in Microsoft 365 Copilot is the rollout of new admin capabilities that let organisations configure, monitor, and control Copilot’s usage more finely. Key additions include:
- Configurable Copilot Search: Administrators can now customise Microsoft 365 Copilot’s search experience for their organisation. This means IT can tailor which sources Copilot draws from and how it surfaces information, enhancing the discoverability of enterprise knowledge. By adjusting Copilot’s search settings to organisational needs, companies can reduce time wasted searching for content and improve decision-making.
- Usage Analytics and Insights: A new usage report in the Microsoft 365 admin centre provides visibility into how Copilot Search is being used. Admins can see total Copilot search queries, average queries per user, usage trends, and even per-user insights. These analytics help identify adoption patterns and potential training needs, ensuring users are leveraging Copilot effectively.
- Cost Controls with Capacity Packs: Microsoft has introduced pre-purchase capacity packs for Copilot’s pay-as-you-go services. Organisations can buy a set number of Copilot messages in advance (monthly or yearly subscriptions) and apply them to Copilot Chat, SharePoint AI agents, etc., avoiding surprise charges. This simplifies budget management and eliminates the need to frequently reset pay-as-you-go setups.
- Retention and Compliance Policies: New settings in Microsoft Purview allow separate retention policies for Copilot and AI-generated content. In practice, this means admins can specify how long Copilot’s outputs or conversational data are retained, aligning with corporate data lifecycle policies. Additionally, Data Loss Prevention (DLP) now extends to Copilot: if content is labelled sensitive, DLP policies can prevent Copilot from processing or revealing that information. This ensures AI won’t accidentally expose confidential data during its responses.
- Custom Dictionary for Terminology: Understanding enterprise-specific jargon is easier now – admins can upload custom dictionaries for Copilot to use in Microsoft Teams meeting transcripts. This improves transcription accuracy for company- or industry-specific terms, so Copilot’s summaries and actions are more precise.
- Governance of AI Agents: For organisations building their own Copilot plugins or agents via Copilot Studio, Microsoft has added a crucial control – admins can block or disable specific Copilot agents that users created. This governance feature (configurable via PowerShell) allows IT to prevent unapproved or rogue AI automations from running in the tenant, maintaining compliance and security standards.
These controls empower IT administrators to deploy Copilot confidently. As one announcement put it, Copilot now provides “centralised tools to tailor search experiences to organisational needs,” improving discoverability and productivity while maintaining oversight. Important compliance safeguards – from retention to DLP – ensure that introducing AI won’t violate company policies or regulations. In short, enterprises can embrace Copilot’s benefits without losing control.
Personalised AI Experiences for Every User
Beyond IT controls, Microsoft is also making Copilot more personalised to each employee, without compromising enterprise security. Traditionally, Copilot draws on organisational data (documents, emails, calendar, etc.) to generate answers. Now, with enhanced personalisation, Copilot can also learn from a user’s communications and preferences (with proper consent) to become an even more helpful assistant.
Enhanced Personalisation Control: A new admin-managed setting lets Copilot use a user’s private work communications, such as Teams chats, Outlook emails, and meeting transcripts, to better serve that user. When enabled by an admin (and opt-in by the user), Copilot will remember details relevant to the individual’s work. For example, it might learn a team member’s nickname or a project’s unique terminology from your chats, then incorporate that context into answers it gives you. Microsoft emphasises that this personalisation data stays private to the user – it’s not shared across the organisation. Users remain in control and can request to disable this at any time, at which point Copilot forgets the personal data. In essence, Copilot becomes “personalised to each user, aiding them in day-to-day tasks by understanding their work on an individual level” while still upholding privacy.
Copilot as a Personal Companion: We’re seeing the concept of Copilot as an AI companion take shape inside enterprises. For instance, Microsoft is expanding Copilot in Viva Engage (the enterprise social platform) to act as a personalised AI aide for employees. This AI companion can proactively surface news or insights tailored to a user’s interests and help them recall important information from across the organisation. By rolling out features like this, Microsoft envisions each employee having an AI that “stays informed, discovers insights, and recalls information” on their behalf – almost like an ever-attentive digital colleague. The result is a more engaging, intuitive experience where the AI understands your role, context, and preferences.
Microsoft is evolving Copilot into a more personalised AI companion for each user. For example, the Copilot in Viva Engage can tailor its feed to keep individuals informed and help them recall insights, acting as a context-aware assistant in the flow of work. Enterprise admins can enable or restrict such personalised features to balance user productivity with compliance.
Benefits of Personalisation: For the enterprise, allowing a degree of personalisation can significantly boost Copilot’s usefulness. It means replies and suggestions from Copilot are more contextually relevant, saving time that might be spent clarifying questions or rewriting generic outputs. Employees get the feeling of a truly “personal” assistant that remembers their needs and style. Microsoft notes that Copilot’s goal is to be “completely personal, built around individual needs, values and expectations… as many different Copilots as there are people using them.”. While that vision must be tempered by organisational policies, the technology is moving in that direction with memory and personalisation features now emerging for business users.
With these new admin and personalisation updates, enterprises can maximise Copilot’s potential. IT admins are equipped with tools to configure the AI’s behaviour, ensure compliance, and manage costs, making Copilot enterprise-ready from a governance standpoint. At the same time, end users can enjoy more tailored, context-aware assistance that adapts to their work patterns and preferences (under the umbrella of admin oversight and privacy controls).
Getting the most from Copilot will require organisations to strike the right balance: enabling the AI to be helpful and personalised, while using the new admin controls to set appropriate boundaries. Companies that succeed in this balance will find Copilot not only boosts productivity but also becomes a trusted, innovative partner in everyday work – a true “co-pilot” for every employee, backed by enterprise-grade control and security.