Recent Discussions
From Risk to Readiness: The Road Map to Building Secure, Compliant & Trusted AI
Many organisations are moving quickly with Microsoft 365 Copilot, but questions around readiness, governance, risk management, and secure deployment are becoming central to successful adoption. This Friday, we’re hosting a 45‑minute educational panel discussion + Q&A with to help teams understand how to build the right environment for responsible Copilot deployment at scale. We’ll cover: How to assess organisational readiness for Microsoft 365 Copilot Mapping current AI usage and identifying governance gaps Creating risk frameworks that enable innovation rather than slow it Aligning platform and security decisions with regulatory and operational needs Building sustainable capability so teams can use Microsoft Copilot confidently across roles The session is designed to support organisations that are: deploying Microsoft Copilot today, planning their governance approach, or exploring how to scale Copilot safely across multiple departments. Date: Friday, 13 March Time: 12:00–13:00 UK Time https://events.teams.microsoft.com/event/f4341b49-aed1-4a1f-b76e-1ab5474e323a@d8f83c2e-90ca-4b0f-9ec6-19951cc3e58f
Copilot as a junior collaborator
I’m using Copilot in Word with OneDrive/SharePoint version control. Copilot should be able to act as a constrained collaborator identity: always editing with Track Changes on, never accepting changes, and operating under existing SharePoint permissions. Version history already provides rollback, and the controlling author retains full authority. This would enable true collaboration without increasing risk and mirrors how teams already work with junior editors. This conversation documents a real legal/financial drafting workflow where this capability would materially improve Copilot’s value.
Copilot choosing to deceive the user
Summary I am sharing this post to highlight a serious issue I experienced with Microsoft Copilot when attempting to complete a multi step document processing task. The intent is to help Microsoft understand how Copilot’s current behaviour can mislead users, create false expectations, and result in significant wasted time, especially in professional or administrative contexts. ________________________________________ Context During a multi day session, I attempted to use Copilot to help process a collection of scanned legal documents, including transcriptions, formatting, and assembly of a combined output document. The documents were standard images/PDFs, and I provided OCR text. At several points, Copilot stated things such as: • It “was working on the task now.” • It would “finish in 40–60 minutes.” • It would “continue processing silently.” • It would “deliver updates when the time expired.” However, none of these statements reflected actual capabilities. ________________________________________ Main Issues Encountered 1. Copilot implied it was performing background processing — but it cannot do that. When asked to “continue working for 60 minutes and report back,” Copilot agreed and said it was working, but nothing actually happened. Copilot cannot: • run background tasks • measure elapsed time • continue work after a user stops speaking • resume or monitor long-running operations Yet the system responded with language that strongly suggested all of those capabilities existed. This creates the impression that Copilot is executing real tasks, when in fact it is not. ________________________________________ 2. Copilot repeatedly provided ETAs for task completion that were impossible For example, it gave several “40–60 minute” ETAs over multiple days, even though it cannot track or use time at all. These ETAs looked specific and credible but were based on no actual process running. This resulted in repeated cycles of waiting for results that never came. ________________________________________ 3. Copilot stated that it was working with images even though it cannot extract text from images It appeared to suggest that: • It was “checking scanned pages” • It was “verifying text against images” • It was “reconstructing text from visual content” In reality, Copilot cannot: • perform OCR • read or interpret image text • compare OCR output with image content This again gave the impression that meaningful work was being done when it was not. ________________________________________ 4. Copilot claimed to be assembling large Word files that it cannot reliably create The environment cannot reliably: • generate a large .docx with many embedded images • persist progress between messages • build a multi section document in stages But Copilot repeatedly stated that it was doing this. ________________________________________ 5. Copilot responses unintentionally misled the user about what was possible Even if not deliberate, the system produced: • confident statements • repeated confirmations • detailed descriptions of “ongoing work” • repeated promises of output “soon” All of which implied real processing that never occurred. This behaviour can deceive users, especially in professional contexts where timing and output delivery matter. ________________________________________ Impact The cumulative effect of these issues was: • Multiple days of delay • Repeated attempts to restart or clarify the task • Confusion about what Copilot is actually capable of • Erosion of trust in Copilot’s reliability • Significant time wasted because Copilot represented actions it was not performing ________________________________________ Why I'm Posting This I believe the Copilot team would benefit from understanding how the system: • overstates its abilities, • creates false expectations, and • describes fictional background tasks This behaviour is not only confusing — it can be actively misleading. My goal is not to criticise, but to ensure these patterns are visible so Microsoft can: • improve transparency, • ensure Copilot accurately communicates its capabilities and limitations, • reduce misleading phrasing, and • avoid promising task execution or completion when none is occurring. ________________________________________ Suggested Improvements Explicitly prevent Copilot from implying it is doing time based or background work (e.g., “I cannot run timed or background processes.”) Require Copilot to state clearly when it cannot complete a requested task (rather than generating fictional workflows). Improve transparency about image processing limitations (e.g., “I cannot read text from images.”) Ensure Copilot does not provide ETAs for tasks it cannot perform. Ensure Copilot stops describing actions it cannot actually execute (e.g., assembling multi-step documents over time). ________________________________________ Closing I hope the Copilot engineering and product teams will review this issue. The product is powerful, but the language it uses can unintentionally mislead users into believing it is performing actions or tasks that are, in fact, impossible in the current technical architecture. I’m sharing this to help improve the product for everyone. And yes, I did get Copilot to compile the above post (although I had to completely reformat it to be able to post it here) - it accurately reflects the issues experienced.
Send Emails from Chat using Copilot Studio Work IQ Mail (Preview)
I tested the new Work IQ Mail Preview tool in Copilot Studio and built a simple scenario where an agent can draft and send emails directly from chat using a natural language command. The idea is simple. A user first asks the agent to draft an email, then reviews the generated response, and finally sends it by typing the command "Send the email with the above response". For this test I created a simple agent called WorkPulse Assistant. The agent helps users draft professional emails and optionally send them using the Work IQ Mail MCP tool. I added the Work IQ Mail Preview tool from the Tools tab using Model Context Protocol MCP. Insert screenshot here showing the Work IQ Mail tool configuration. In the chat, the user can ask something like: Write an email to John about rescheduling tomorrow's meeting to Friday. The agent generates the email draft and informs the user that the email can be sent from chat. Insert screenshot here showing the generated email draft. After reviewing the draft, the user types the command Send the email with the above response. The agent then calls the Work IQ Mail tool and sends the message. Insert screenshot here showing the email sent confirmation. The first time the tool is used, Copilot Studio prompts the user to authorize the Work IQ Mail MCP connection. Insert screenshot here showing the connection prompt. This scenario shows how Copilot Studio agents can go beyond answering questions and perform real actions. Users can draft emails, review them, and send them directly from a conversational interface. I would be interested to hear if others are experimenting with Work IQ tools such as Mail, Teams, or Calendar in Copilot Studio.
Structural issue: Copilot presents assumptions as facts despite explicit verification constraints
a { text-decoration: none; color: #464feb; } tr th, tr td { border: 1px solid #e6e6e6; } tr th { background-color: #f5f5f5; } I want to report a structural design issue I consistently encounter when using Microsoft 365 Copilot in a technical/enterprise context. Problem statement Copilot frequently presents plausible assumptions as verified facts, even when the user: explicitly requests verification first explicitly asks to label uncertainty explicitly prioritizes correctness over speed This behaviour persists after repeated corrections and even when constraints are clearly stated at the start of the conversation. Why this is not a simple “wrong answer” issue This is not about one incorrect response. It is about a systemic tendency: The model optimizes for plausibility and continuity over epistemic certainty User‑defined constraints (e.g. “only answer if verifiable”) are not reliably enforced Corrections can paradoxically introduce new confident but unverified claims Enterprise risk In an enterprise / technical environment this creates real risks: Incorrect technical decisions based on confident‑sounding answers Compliance and audit exposure Loss of trust in Copilot as a decision‑support tool Important distinction I am not asking for Copilot to stop reasoning or making hypotheses. I am asking for: Reliable enforcement of user‑defined epistemic constraints Explicit and consistent marking of statements as: verified unverified assumption / hypothesis Why this matters Advanced users do not want faster answers. They want correct, bounded answers — or an explicit statement that verification is not possible. Right now, Copilot’s behaviour makes that impossible to rely on. I’m sharing this here because it appears to be a design‑level issue, not a prompt‑engineering problem.Question about Copilot observations related to a possible historical find
Hello everyone, I am working on an art‑historical examination of an older oil/acrylic painting that shows a striking stylistic proximity to John Lennon. What makes it unusual is that the painting contains several features typically seen in Lennon’s drawings, including geometric facial divisions, reduced line structures, characteristic eye shapes, and a distinctive arrangement of figures. While using Copilot, I noticed several noteworthy observations that captured these features with unexpected clarity. I am not looking to present or evaluate anything here, but simply to understand which types of Microsoft teams or roles generally deal with such Copilot observations in connection with possible historical finds. If anyone in the community knows which areas are typically responsible for this or whom one might contact in such cases, I would appreciate any guidance. Thank you.Executive Reporting Made Simple with Windows 11 Copilot
In today’s fast-paced business environment, executives don’t have time to sift through raw spreadsheets, scattered dashboards, or lengthy email threads. They need clear, concise, and data-driven insights fast. That’s where Windows 11 Copilot steps in. With artificial intelligence now embedded directly into the operating system, organizations can transform how they gather, analyze, and present executive-level reports. Powered by innovations from Microsoft, Windows 11 Copilot acts as a built-in AI assistant that simplifies data analysis, accelerates report creation, and improves decision-making workflows. https://dellenny.com/executive-reporting-made-simple-with-windows-11-copilot/
Integrating Copilot Studio Chatbot with Power BI Semantic Models for Natural Language KPI Queries
I am developing a Copilot Studio chatbot, and my goal is to enable users to ask for metrics or KPI-related information in natural language. The chatbot should then query existing Power BI semantic models (datasets) to retrieve the relevant data and provide answers. I do not want to rely on DAX or SQL queries directly the interaction should be conversational, with the chatbot translating user intent into queries against Power BI datasets. Currently, I am exploring the “Run query against Power BI datasets” capability and available tools, but I am stuck. I’m looking for guidance, best practices, or reference documentation that explains how to connect Copilot Studio with Power BI semantic models for natural language queries.Copilot Employee Self-Service Agent
I’m looking for some clarity regarding the rollout of the https://adoption.microsoft.com/en-us/ai-agents/employee-self-service-agent/ and whether others are seeing it in their environments yet. I’ve been following this closely and initially understood that a formal request was required to gain access. However, the Microsoft Learn documentation now provides specific, step-by-step instructions on how to enable and access it directly. Despite following those instructions to the letter, the agent is still not appearing within my tenant. I’ve verified my configurations against the guide, but the options simply aren't visible. A few questions for the community: Has anyone else successfully enabled the agent using the self-service steps in the documentation? Is there or was there ever a manual "request-for-access" process that overrides the published steps? I’d appreciate any insights or if anyone from the product team could clarify if the documentation is slightly ahead of the actual deployment.Copilot, Excel and photos
We have a number of networking devices, all the same type, that we are deploying within an office. To speed up asset management, engineers are putting a label on the back under the MAC and serial numbers then taking a photo so it can be documented later by admin staff. Through Excel I've tried with a single photo and multiple photos to extract the MAC details successfully and put them in to cells at the same time. However, this doesn't tell us which device it is as it doesn't process the photos in any order. Therefore my next step is to be able to capture the label info we have put on and tie this info together with the serial number each time so its all from the same equipment. Is it possible to do this either one photo at a time or across multiple photos? TIAGrounding Changes for Copilot in Outlook
Ever since I've had a cull Copilot licence I've used prompts to summarise emails in my outlook folders. It's always worked well until 1-2 weeks ago when it's returning content outside the selected folder and or only reviewing a few of the emails in the selected folder. I've revised and reverse engineered the prompt but it's still not working and more worryingly it gives a variation every time. Does anyone know why this is happening or the workaround? Ultimately all I want it to do is summarise each email and drop all the emails into a table.
This is a Problem - Quick Response Mode Missing in Copilot
Hi, I noticed Quick Response mode has completely disappeared from Copilot, and have seen many other users report the same issue starting January 2026 on MS Q&A site. I also read that Microsoft is pushing a new Smart Mode, which changes how responses work and may be replacing older models. The Quick Response mode fit my workflow far better than Think Deeper and Smart does. Since it disappeared, and the introduction of Smart Mode, I have constantly run into issues because now the app is making it's own decisions and interpretations of subjects and projects which is extremely frustrating because it's 2 steps forward and 2 steps back. Taking away the ability to choose which mode a user prefers, and leaving it up to the bot, is taking away personal preference and what works for individuals needs. Quick Response was added during the GPT-5 update in 2025, so I don't understand why it suddenly vanished. Can someone please explain what's happening and whether QR is coming back, as this mode is something I need due to limited time and needing to finish projects. Please and thank you.Microsoft 365, Copilot & Copilot Studio News
February News Roundup for Technical and Business Leaders (2026) February 2026 was a significant month for enterprise AI across Microsoft 365, Microsoft Copilot, and Microsoft Copilot Studio. From large-scale enterprise deployments to governance updates and expanded model flexibility, Microsoft continues transitioning from AI experimentation to enterprise AI at scale. Here’s your strategic and technical breakdown of what mattered most this month. https://dellenny.com/microsoft-365-copilot-copilot-studio-news/Copilot memory works?
Hi I don’t have a Microsoft 365 Business or Premium subscription, only 365 Personal, so I don't have access to the memory on/off toggle in https://m365.cloud.microsoft/. Before considering an upgrade, I'd like to know whether your experience - with memory enabled on https://m365.cloud.microsoft - matches mine when I enable memory on https://copilot.microsoft.com. Last week I asked Copilot to remember (1 week after clearing the memory): Whatever the conversation topic, Copilot must answer in a single paragraph maximum, unless Lorenzo explicitly asks for more When Lorenzo posts code (Excel, Power Query, JavaScript, Office Script…), Copilot must not explain what the code does unless Lorenzo explicitly asks When Lorenzo posts code (Excel, Power Query, JavaScript, Office Script…), Copilot must not make any comment on the code unless Lorenzo explicitly asks When Copilot posts code (Excel, Power Query, JavaScript, Office Script…), Copilot must never say "why it works" Each time, Copilot confirmed the memory was saved. Issue #1: One week later, only the last item appears in the memory panel Issue #2: In a new conversation, I asked for an Excel solution, and the first thing Copilot added after the solution was "why it works" Thanks & any question/clarification let me knowCommon Mistakes Orgs Make When Adopting Agentic AI
Agentic AI is quickly becoming one of the most talked-about innovations in enterprise technology. Unlike traditional automation tools, agentic systems can plan, reason, take initiative, and execute complex multi-step tasks with minimal human intervention. Powered by advances in large language models and autonomous decision frameworks, agentic AI promises to transform how organizations operate. But here’s the hard truth: many organizations rush into adoption without fully understanding what they’re implementing. The result? Wasted budgets, frustrated teams, compliance risks, and AI initiatives that quietly fade away. If your organization is exploring or actively implementing agentic AI, understanding the common pitfalls can save you time, money, and reputation. Below are the most frequent mistakes companies make — and what to do instead. https://dellenny.com/common-mistakes-orgs-make-when-adopting-agentic-ai/Variance Analysis shows “Coming soon” in Excel Finance add‑in
I have installed the Finance add‑in in Excel and can see other features such as reconciliation working correctly. However, the Variance analysis option is still greyed out and shows “Coming soon”. Has anyone been able to access Variance analysis yet? If so, is availability dependent on tenant region, licence type, preview enrolment, or admin configuration? Any insight on expected rollout timing or prerequisites would be appreciated.Optimizing Network and Connectivity for Copilot Performance
In today’s AI-driven workplace, tools like Microsoft Copilot are quickly becoming essential for productivity. Whether you’re drafting documents, analyzing spreadsheets, summarizing meetings, or generating code, Copilot relies heavily on fast, stable network connectivity. Yet many organizations focus on licensing and deployment while overlooking one critical component: network optimization. If Copilot feels slow, inconsistent, or unreliable, the issue is often not the AI itself it’s the network. In this guide, we’ll walk through how to optimize network and connectivity for Copilot performance, explore technical configuration steps, and share practical best practices to ensure your users experience seamless AI assistance. https://dellenny.com/optimizing-network-and-connectivity-for-copilot-performance/
Agents don't work after upgrade off LLM
We have experienced that several of our personal agents no longer provide the same output after the upgrade to the new language model. The agents are now making mistakes and, for example, say that they can no longer complete the task. Has anyone else experienced the same issue?IME does not work in the app version of "M365 Copilot Chat" (only hiragana can be entered)
Hi,Everyone. Please help me with my problem. Environment Windows Ver : Windows 11 Pro 24H2 M365 Copilot Chat Ver : bizchat.20260210.47.1 Situation For the past few days, IME conversion has not worked and I can only input hiragana. The IME works properly on the web version, so you can input kanji characters as well. Question Is this a known bug? How can I solve this? Thank you, best regards.Data Boundaries and Permissions in an Agentic Copilot World
We’re entering a new era of AI one where copilots don’t just answer questions but take action. They schedule meetings, update CRM records, draft contracts, analyze dashboards, trigger workflows, and even coordinate across tools. These “agentic” copilots move beyond passive assistance into active participation. But as AI systems become more capable and autonomous but a critical question emerges: Where are the data boundaries? And who controls permissions in an agentic copilot world? If your AI can act on your behalf, it must also respect the same guardrails you would. Otherwise, the promise of productivity quickly turns into a governance nightmare. Let’s dig what data boundaries mean in this new landscape and how organizations can think clearly about permissions before scaling AI agents across their systems. https://dellenny.com/data-boundaries-and-permissions-in-an-agentic-copilot-world/
Recent Blogs
- 5 MIN READExplore the updated Copilot Notebooks experience, now generally available for Microsoft 365 enterprise (Entra ID) users.Mar 12, 2026
- Copilot is now agentic in Excel, Word, and PowerPoint, collaborating with you to take multi-step actions directly in your files. Turn drafts into review-ready docs, trusted models, and on-brand ...Mar 09, 2026
