Site index
All routes. One list.
Human taxonomy of the public site. Machine sitemap stays at sitemap-index.xml. LLM entry points stay at llms.txt and llms-full.md.
Primary
Core pages
Home Kinetic intro, help blocks, revolving panes, Wave 0 Framework Four arcs, twelve panes, operating rhythm Start · Wave 0 Security → GRC → Cost & Value → Licence → Adoption Enterprise AI Hub: pains, matrix, offer ladder, CTA Four pains Spending, exposure, value, fragmentation Operating framework Establish→improve cycle + eight domains Board view Board questions, 90-day decisions, scorecard Counsel Six members, dials, session example Offerings Foundation, Operate, Scale, Pro bono, Sector 3 Personas Decision archetypes + survey to package mix Roadmap builder Interactive 0–30–60–100 timeline About Authorship, method, contact Ask You ask, we answer — Copilot & agents Q&A Blog Field notes and SEO articles Microsoft AI Copilot, agents, Frontier, CCS, Foundry guides Roadmap intel MRC Copilot/Agent roadmap pages Podcast Audio stub · Gemini TTS when connected 12 Panes TV Explainers, walkthroughs, Counsel sessions as film
Counsel · members
Six roles
Taxonomy · panes
01–12
Brand and ops
Studio pages
Machine
LLM and crawlers
Ask & blog
You ask, we answer (82)
- If we remove someone's access in ServiceNow, how long before Copilot stops showing it?
- Copilot now has legal plugins. Should our lawyers be using them?
- Should Teams announce out loud when recording or transcription starts?
- People are sharing Copilot chats by link. Who can see them, and should we turn it off?
- Staff can't upload Confidential files to our Copilot Studio agent. Is that about to change, and should we be pleased?
- Does our Copilot DLP policy cover Cowork, or do we need a second one?
- Can we see agent reporting in the Agent 365 Dashboard without buying Agent 365?
- We’ve paid for Copilot connectors to ServiceNow and Jira — how do we prove anyone uses them?
- Our agents need Dataverse access — do we give them a service account or an Entra agent user?
- Edge Watermarking is coming — does a ‘Confidential’ overlay actually stop exfiltration?
- Facilitator will warn about meeting scams — is that the same as Teams phishing detection, and should Security rely on it?
- Copilot Search is coming to SharePoint and OneDrive — does that amplify oversharing before we’ve finished Wave 0?
- Copilot can now cite the memories that shaped an answer — what should Privacy tell staff before we leave Memory on?
- Teams now scores third-party apps and agents against our trust criteria — do we still need a human approval gate?
- Teams is adding AI phishing detection for risky chats — does that replace Safe Links and Defender for Office 365?
- Outlook Copilot Chat is coming to shared and delegate mailboxes — who sees what before we roll it to EAs?
- Purview DLP admin units are coming for Copilot and Cowork — can BU security own policies without tenant-wide rights?
- SharePoint’s Agent chat web part will put any M365 Copilot agent on a modern page — how do we stop intranet sprawl?
- Purview network DLP via Entra GSA will cover all websites — do we finally close the ChatGPT paste gap?
- Sales Development agent can write to Salesforce via MCP — do we let it qualify leads on its own?
- Work IQ will let agents upload files into SharePoint and OneDrive — how do we stop that becoming silent sprawl?
- Will our old Teams-and-Copilot retention policy stop covering Copilot prompts?
- Federated Copilot connectors can write to ServiceNow and CRM — how do we stop that becoming a free-for-all?
- We need Copilot proofreading in Word with Track Changes — and one shared prompt for the whole org. Is that possible?
- Copilot in Excel edits by default — can we make it plan first on finance workbooks?
- Will the Workforce Insights agent let managers ask Copilot about their staff?
- Will Microsoft 365 Copilot now show everyone our Dataverse and CRM data?
- Can Teams stop a deepfake ‘CFO’ on a call before someone pays?
- Will premium models in Copilot Chat run up a bill we didn’t plan for?
- Should agents pause for a human before sending mail or closing tickets?
- How do makers see agent credit cost before Finance sees the invoice?
- Should we standardise agents with a Skills catalog — or is that more sprawl?
- Should makers clear Agent Readiness before we let them publish?
- Should Copilot Studio agents get their own mailbox and Teams identity?
- Can Copilot Studio agents ‘self-learn’ — and who approves the changes?
- Can we export Agent Dashboard metrics without creating a privacy incident?
- Should we enrol everyone in Copilot Frontier for Excel Agent Mode?
- Who should publish organisational Copilot prompts — and can we delegate?
- How do we stop agents using the wrong MCP tools?
- Where do Copilot Pages and Notebooks live — and can we govern them?
- Does Purview Copilot DLP cover Copilot Studio agents?
- Does turning off agent Sharing stop Agent Builder sprawl?
- What's the difference between Available to and Deployed to?
- Do Copilot Studio agents always need admin approval?
- Must we allow Move data across regions for Copilot Studio?
- Why can’t Facilitator join our Teams channel meetings?
- Why do Researcher and Analyst ignore our agent settings?
- Should we turn on ‘AI experiences enabled by usage-based billing’?
- Do new Copilot Studio agents get Entra Agent IDs automatically?
- How do spending policies stop runaway Copilot Credits?
- Do we need a Conditional Access policy just for Microsoft Copilot?
- Should we pin agents for everyone in Microsoft Copilot?
- Should we let users connect their own Jira and Confluence to Copilot?
- Why does Copilot Studio block Knowledge uploads with a data policy error?
- Should we route Copilot Studio makers out of the Default environment?
- How do we see which agents are reading our SharePoint sites?
- Why does Agents → All agents suddenly show a hundred entries?
- Can we set monthly Copilot Credit caps per agent?
- Does a Microsoft 365 Copilot seat cover GitHub Copilot harness agents?
- Should we allow ChatGPT or Claude connectors into Outlook?
- Should we block maker credentials on Copilot Studio agents?
- Can we run Copilot Studio agents from the default environment?
- Do we need Agent 365 if we already bought Microsoft 365 Copilot?
- How do we stop makers publishing anonymous Copilot Studio agents?
- Should we block Direct Line and external Copilot Studio channels by default?
- Should we allow Anthropic Claude models inside Copilot?
- Are Copilot Graph connectors safe to turn on?
- Can we audit and retain Copilot prompts and responses?
- Can we block Copilot agents by default until IT approves?
- What does ‘prepare our org for agents’ mean for IAM?
- How do we stop everyone creating Copilot Studio agents?
- Do Copilot Studio credits burn when staff already have Copilot seats?
- Can we turn off Copilot’s web search grounding?
- Restricted SharePoint Search is retiring — what do we do?
- What do we do with dormant Copilot seats?
- Can guests use our Copilot agents?
- Do sensitivity labels stop Copilot reading files?
- What do we do about ownerless agents?
- Should we turn on pay-as-you-go for SharePoint agents?
- How much permission should an agent get?
- Who actually needs a Microsoft Copilot licence?
- Must we fix oversharing before turning on Copilot?
Field notes (77)
- A 12-pane maturity radar
- A board pack rhythm for AI
- After Wave 0: Move and Grow
- Agent registry, privilege ratings, and SharePoint agent control
- Bounded frontier work
- Build a champion network
- Calm enterprise voice
- CCS: last-accessed retention and SharePoint citation analytics
- Clinics that convert to sustained use
- Copilot Agents: extend Copilot without losing the plot
- Copilot Control System (CCS): the admin spine
- Copilot Cowork: agentic tasks with human approval
- Cost, value, and use it or lose it
- Counsel dials for a cost-out programme
- Counsel dials for a design-led scaleup
- Counsel dials for a regulated bank
- Deployer versus provider duties
- DPIA triggers for AI systems
- Every pane has insight and action
- Extend IR for AI misuse
- Finding dormant seats early
- FinOps habits for AI spend
- Flexibility under disruption
- From pilot purgatory to portfolio control
- Frontier features and Frontier agents
- Graph Copilot usage API — fuel for the Steward pane
- GRC that enables work
- Harden before broad Copilot: security that owns AI risk
- ISO 42001 as a standards bridge
- Keep a horizon watchlist
- Least privilege for makers and agents
- License proven cohorts
- Literacy that changes behaviour
- Measuring benefit without vanity metrics
- Meet the 12 Panes Counsel: judgment inside the window
- Meeting the Counsel
- Microsoft 365 Copilot: work grounding without theatre
- Microsoft Agent 365: observe, govern, secure
- Microsoft Copilot: the platform view
- Microsoft Foundry: agents, models, and tools in one plane
- Microsoft Scout: desktop agent with SharePoint in the loop
- NIST AI RMF through the window
- O*NET, ESCO, and SFIA together
- One clear organisational story
- OneDrive agents: personal knowledge, shared risk
- Outlook Prioritize by Copilot — and citations you can check
- Potential is not realised value: prove AI beyond usage theatre
- Publish a short AI value case
- Reading adoption signals honestly
- Redesigning work with SFIA ESCO O*NET
- Retrospectives tied to panes
- Security and oversharing: the amplification problem
- See before you scale
- See before you scale: thin awareness beside Wave 0
- SFIA as the accountability bridge
- Stand up an AI and agent register
- Standards bridges, not badges
- Start with a shadow AI inventory
- Stop spraying seats
- Talk about Wave 0
- Task maps for priority roles
- The 12 Panes operating rhythm
- The Cartographer: what we know
- The Chair: right order, clear close
- The fully loaded AI cost stack
- The Guardian: what could go wrong
- The messy middle of AI seats
- The Pathfinder: what else to try
- The Pulse: how people feel
- The Steward: where value sits
- Use it or lose it: licence doctrine for AI seats
- Voice chat in M365 Copilot — and the Teams Stop button
- Wave 0 as a 100-day path
- Wave 0: the steady base that ends pilot purgatory
- What changes for your people
- Why the window is the motif
- Why we built a window not another maturity model
DIGITAL SKILL WORKS LTD / 12 PANES