AI Features
How Muin's AI works across documents, forms, and finance — what it automates, how accurate it is, and how your data stays private, never training public models.
49 questions
How does the AI keep my data private?
Muin takes your data privacy seriously. Here's how we protect your information: **Enterprise-Grade AI Infrastructure:** Muin's AI runs on AWS Bedrock: AWS's managed foundation-model service for enterprise AI workloads. Per AWS's published data privacy commitments: - Your data is **never used to train AI models** - AI request content is **deleted within 30 days** and kept only for error recovery - Complete isolation between customers **Encryption:** - All data encrypted in transit (TLS) - All data encrypted at rest (AES-256) - Field-level encryption for sensitive PII (AES-256-GCM) **Compliance:** - SOC 2 Type II: Planned (architecture designed for compliance) - HIPAA: Not supported. Muin is not HIPAA compliant and does not sign BAAs - GDPR: Designed for compliance - Full audit logging **Your Control:** - You can delete your data at any time - Data is never sold or shared for advertising. Processing occurs through vetted sub-processors listed in our [DPA](/legal/dpa) - No training on your documents without explicit consent
Why did Muin categorize my document incorrectly?
AI categorization is based on analyzing document content, but it's not perfect. Here's what to do: **To correct the category:** 1. Open the document 2. Click **Edit Details** or the category badge 3. Select the correct document type 4. Save your changes **Why this happens:** - Unusual document layouts - Mixed content types (e.g., invoice with contract terms) - Low-quality scans - Documents in uncommon formats **Your corrections help!** When you correct a categorization, Muin learns from it. Over time, similar documents will be categorized more accurately. **Consistently problematic?** If certain document types are always miscategorized, contact support. We may be able to improve our models for your use case.
What can I ask the AI assistant?
Ask Muin can help you with a wide range of tasks: **Finding Information:** - "Show me invoices from last month" - "Find contracts expiring in 30 days" - "What policies need acknowledgment?" **Understanding Documents:** - "Summarize this contract" - "What are the key terms in this agreement?" - "Extract payment terms from this invoice" **Getting Help:** - "How do I upload a document?" - "What does this error mean?" - "Help me set up a new employee" **Taking Action:** - "Create a task to review this document" - "Send this invoice for approval" - "Remind me about contract renewal in 30 days" **Tips for better results:** - Be specific: "Invoices over $1000 from Acme Corp" vs just "invoices" - Muin knows your current page context - Ask follow-up questions naturally How-to answers come from the Help Center, in the menu names you see. Ask Muin explains how to use a feature, not how a score, rule or price is worked out behind the scenes. If it is not sure, it says so and passes your question to the Muin support team.
What does the Usage Analytics dashboard show?
The Analytics dashboard (**Analytics** in the sidebar, under Intelligence) gives owners and admins a workspace-wide view of how Muin is being used: - **Agent performance**: runs, success rate, and average completion time for AI agents - **Workflow completion**: how many automations ran and finished successfully - **Token usage**: LLM tokens consumed across all agent runs - **Feature adoption**: which modules and features your team actually uses Use the period selector (7, 30, 90, or 365 days) to compare activity over time.
Why is my Analytics dashboard empty?
Analytics are computed from your workspace's actual activity, so a new workspace starts empty. Data appears as your team: 1. Runs AI agents or chats with the assistant 2. Executes workflows 3. Uses platform features day-to-day If you expect data but see none, check the period selector: a 7-day window on a quiet week can be empty while the 90-day view has plenty. Data refreshes automatically; there is nothing to configure.
Who can see the Usage Analytics dashboard?
The Analytics dashboard is limited to **owners and admins**: it shows workspace-wide operational metrics that aren't relevant (or appropriate) for every member role. If a team member needs analytics access, an owner can change their role in **Settings** → **Team**. For business reporting that anyone can run (financial statements, vendor spend, donor reports), use the **Reports** module instead.
Can I talk to Muin instead of typing?
Yes: in **Chat**, click the microphone button next to the message box, speak, and stop the recording. Your words are transcribed automatically and the assistant responds as if you'd typed them. Short, action-shaped phrases (like dictating an expense) are recognized as quick commands; everything else flows as normal conversation. Muin auto-detects which you meant.
Does voice input work in languages other than English?
Yes: transcription supports multiple languages, so speak in the language you work in. For best accuracy in any language: minimize background noise, speak naturally, and keep clips to a sentence or two. You can always edit the transcription before sending if a name or organization-specific term came through imperfectly.
What's the difference between voice input and the voice briefing?
They're opposite directions. **Voice input** is you talking to Muin: recording in chat and having it transcribed. The **voice briefing** is Muin talking to you: your daily briefing narrated as a podcast-style audio episode you can play in-app or subscribe to via a private RSS feed. Voice input lives in **Chat**; the voice briefing lives at **Intelligence** → **Daily Briefing** under Preferences.
What happens when I upload an invoice?
Muin's AI reads it automatically, vendor details, invoice number, dates, line items, tax, total, and payment terms, and files them as structured data with a **confidence score**. No button to press; extraction is part of document processing. Review the result on the invoice detail page: AI-extracted values are marked, and you can edit any field to correct them. Duplicates (same vendor + invoice number) are flagged before they become double payments.
The AI extracted a wrong value from my invoice: what do I do?
Edit the field on the invoice detail page: your correction takes precedence over the AI's reading everywhere downstream (approvals, reports, exports). Start with low-confidence fields: the confidence score tells you where to look first. Totals, dates, and identifiers deserve a glance even at high confidence, since they drive payments.
What is Business Memory and where do its entries come from?
Business Memory is the institutional knowledge Muin learns about your organization, facts, decisions, outcomes, patterns, and relationships, extracted automatically from chats, documents, workflows, and agent runs, or added manually with **Add Memory**. The AI draws on this memory for context, which is how it can answer questions like "what are Acme's payment terms?" with *your* organization's actual answer. Owners and admins manage it at **Intelligence** → **Memory**.
Muin remembered something wrong: how do I fix it?
Open **Intelligence** → **Memory**, find the entry (filters and search help), and mark it **Disputed**: disputed memories are held back from AI context. If the fact changed rather than being wrong, also click **Add Memory** to record the current version. Entries that should never have been captured can be deleted outright. A quick weekly pass over **Auto-extracted** entries keeps memory trustworthy.
What does verifying a memory actually do?
Verified memories carry more weight when the AI assembles context for answers, agents, and sidebars, and disputed ones are silenced. Verification records that a person confirmed the entry, so downstream AI behavior rests on knowledge someone vouched for. Auto-extracted entries are usually right, but the consequential ones (money, terms, deadlines, commitments) deserve a quick check against reality before you rely on them.
How do I create a reminder by just typing what I want?
On the **Reminders** page, use the **AI Reminder** box: type something like *"Remind me to call the Akram Foundation in 3 days at 2pm"* and Muin parses the timing and title for you. The same works conversationally in Chat. If your phrasing can't be parsed, the quick-add form right below takes a title, date, and time, and **More options** opens recurrence, priority, and record linking.
Can reminders repeat automatically?
Yes: open **More options** when creating (or edit an existing reminder) and set the recurrence: daily, weekly, or monthly. Completing one occurrence automatically schedules the next. Recurring reminders fit rhythms like payroll submission or monthly grant-report checks: set once, and the next nudge always exists.
What happens when I snooze a reminder?
Snoozing reschedules the nudge to a later moment, presets like an hour from now or tomorrow morning, without marking the work done. The reminder shows in the **Snoozed** count until it fires again. Snooze is for "not now"; **Complete** is for "done". The summary cards at the top of the Reminders page keep the difference visible.
What is the Chief of Staff page?
**Sidebar → Intelligence → Chief of Staff** is one short list of the few things that need you today, ranked, each with the action that clears it. Below it, **Everything else, in order** shows what is coming without asking you to act. It replaces opening six module dashboards to find out what changed. Four companion pages sit in the link strip across the top of it: | Page | Answers | |---|---| | **Settings** | What the assistant may do on its own | | **Readiness** | Which of an inspector's questions you cannot answer yet | | **What if** | What happens to everything else if you change one thing | | **Board motions** | Whether what the board decided was actually done | **Coverage** has its own sidebar row, directly under Chief of Staff. Each page carries its own permission, so your sidebar may show fewer rows than a colleague's. When nothing needs you, the page says "Nothing needs your attention right now" rather than inventing work.
Can the assistant act on my behalf, and what stops it going too far?
Only if somebody grants it a **mandate**, and the mandate is where the limits live. Open **Chief of Staff → Settings** and click **Let it act for someone**. A mandate names six things: **Who it acts for**, **Which assistant**, **What it may read**, an **Amount ceiling**, **Actions a day**, and an **Ends on** date. Two things worth being clear about: - **An amount ceiling is a limit on the assistant, not a replacement for your approval rules.** Money still goes through the normal approval chain. If you leave the ceiling blank, the card says so. - **Revoking is auditable.** Click **Revoke this mandate** and confirm; the mandate stays on the page showing the date its authority ended, so the revocation itself can be checked later. Granting a mandate needs permission to manage assistant mandates.
Who handles my approvals while I'm on leave?
Whoever you set up as **coverage**. Check it at **Sidebar → Intelligence → Coverage**, which tells you **Covering you**, **If they do not act**, the **Window**, **What is covered**, and **Stays with you above** - the amount over which decisions wait for you regardless. Coverage is read-only on that page. You set it up in your Inbox: click **Set up coverage**, or go to **Sidebar → Inbox** and use the **Delegate** drawer. If nobody is covering you the page says **Nobody is covering you** and offers the link, which is the answer you want to see *before* you travel rather than after. Delegations apply to new tasks from the moment you set them, not retroactively.
How do I know if we're ready for an audit or an inspection?
**Chief of Staff → Readiness** asks the questions an inspector would and answers them from your own records. 1. Choose **Who do you want to be ready for?** - **Tax authority**, **Fire marshal**, **Single audit** or **Insurer** 2. Read the three counters: **Answered**, **Gaps**, **Does not apply** 3. **What you cannot answer yet** lists the gaps; **Everything else they would ask** lists what is already answered, each with the evidence behind it 4. On a gap, create a task so somebody owns closing it - it lands on your Inbox list titled "Close the readiness gap: <question>" 5. **Ask again** re-runs the questionnaire after you have filed the missing records Answers come from your records, not from a template. A question with no record behind it is reported as a **gap**, never answered optimistically. This page needs permission to view risk incidents.
Can I build my own AI agent?
Yes. **Sidebar → Agents → My agents → Create Custom Agent** asks one question: **What should this agent do?** One or two sentences is enough. Click **Draft it** and Muin drafts the name, instructions, tools, schedule and what each finding includes. Owner and admin roles only. The review screen shows each part so you can check it. **Ask to change** lets you ask for a change in plain words and accept it or undo it. **Try it** runs the draft on your records from the last 30 days and shows what it would have found; nothing is saved and nothing changes. **What it costs** shows the credits per run and about how many credits a month it will use. Then click **Turn on**. Tools come from the same catalog Ask Muin uses, filtered to what you are allowed to use: **What it can look at**, and **What it can do (asks you first)**. A tool that changes a record never runs on the agent's word: each change it proposes waits for a person to approve it in **Inbox**. You choose how careful it is: **Fast**, **Balanced** or **Smart**. Fast is cheapest, Smart is most careful. Saving a new version keeps the old one, and the **Versions** tab lets you **Roll back**. **Advanced** opens the full seven-step wizard with tunable settings.
Does building an agent cost credits?
Yes. Drafting an agent, **Ask to change**, the problem check, test runs and replays are AI calls, labelled **Building and testing** on **Settings → Billing → Usage**. Each month your plan includes building and testing credits, spent first and shown as "(included)": 100 on the Free trial and Starter, 500 on Growth, Scale and Enterprise. After those, building and testing uses your monthly AI credits like any other AI work. When both are used up, the builder shows a card to add credits. Your draft is kept, and the step you were on runs again once credits are added.
How do I make an agent run on a schedule?
Open the agent and go to its **Schedule** tab. Pick from **Schedule**: - **Not scheduled - runs when something triggers it** - **Every hour** - **Every day at 6:00am** - **Every Monday at 6:00am** - **The 1st of every month, 6:00am** - **Custom schedule…**, which gives you a **Cron expression** field (for example `0 6 * * 1-5` for 6am on weekdays) Click **Save schedule**. The card then shows the schedule as a badge, or **Not scheduled**. Two things to know: - You set the schedule **after** the agent is on, not during setup. Setup shows the schedule the agent ships with, read-only. - Under **When something happens** on the same tab you can also run it on an event: pick it from **Run this agent when**, and click **Add trigger**. If you see "This agent has no activation yet, so it cannot be scheduled", you are looking at a library definition rather than a copy you turned on.
How do I see what an agent actually did, and what it cost?
Open the agent, use its **Runs** tab, and click a run. The run page shows the **Findings** it produced and the inbox tasks it raised, the **Extracted Fields** it read, the **Execution Steps** and **AI Decisions** it worked through, any **Quality Flags**, and - when it failed - the error code, message and detail. On cost: a run records **input, output and total token counts**, the model used, the confidence and the duration. **It does not record a dollar figure.** An agent that worked from your records rather than a document reports "Not used. Worked from your records." instead of a model and token count. Two notices worth knowing: - A run queued for more than about a minute and a half says **Not started yet**, and states plainly that nothing has run and nothing has been charged. - A run whose document is missing says **No document to read** and tells you retrying will not help. A failed run offers **Retry**; one still going offers **Cancel**.
Can Ask Muin tell me how to use Muin itself?
Yes. Ask a "how do I" question and Ask Muin searches the Help Center rather than answering from memory. Ask Muin also knows Muin's menus, pages, tabs and links, so it can tell you how to reach a page even when it has no help guide yet, for example "Giving > Home > More tools > Pledges, then the Installment Pledges tab". You get up to five results, each a clickable link to the full guide plus a short summary. A result labelled **Doc** is a guide; **FAQ** is a short answer. Clicking one opens the article in the in-app Help Center. Help content is the same for everyone - it is not your business data - so no module permission is needed to search it. Any signed-in user can. If the feature is not in your organization's Plan, Ask Muin says so instead of walking you through something you cannot open. If your role cannot open it, Ask Muin tells you and offers to request access. If nothing in the Help Center answers the question with confidence, Ask Muin says "I'm not sure" and passes your question to the Muin support team instead of guessing. Open it with **Ask Muin** in the module bar at the top of any page, or press **Cmd+J**.
Why does Ask Muin say a feature is not in my Plan?
Before Ask Muin explains how to do something, it checks whether your organization's Plan includes that feature. If it does not, Ask Muin says: "That isn't part of your Plan. Your admin can check what your organization has in Settings." It also names the page to look at: **Settings > Billing**. This saves you from following steps for a page you cannot open. If your organization has the feature but your role cannot open it, the reply is different: "You don't have access to ... yet. Ask your admin, or request access below." Use the request button under the reply. If you think the answer is wrong, ask your admin to check **Settings > Billing**.
Why did Ask Muin say I'm not sure?
Ask Muin answers how-to questions from the Help Center. When nothing there answers your question with confidence, it says so rather than guessing: "I'm not sure about that one yet. I've passed your question to our team, and you'll hear back soon." That means your question went to the Muin support team. You do not need to file a ticket yourself. Questions the Help Center cannot answer also tell us which guides to write next. If Ask Muin says "Please try asking another way, or browse the Help Center" instead, the question could not be passed on. Try different words, or use **Contact Support** on the Help Center page.
Why won't Ask Muin explain how a score or rule works?
Ask Muin explains how to use Muin: how to do something, where to find it, whether you can, and what a page is for. It does not explain how the product works behind the scenes, such as: - how a score, rating or risk level is calculated - the rules and limits that decide an outcome - how prices, fees or totals are worked out - how the AI itself is set up Ask one of these and Ask Muin replies: "I can't explain how that works behind the scenes. Here is the help article for this feature." The article covers what the feature does for you and how to use it.
Why does the mobile app open my browser, and do I have to sign in again?
No, you do not have to sign in again. Some things are easier on a big screen - signing a document, a long form, a wide report - so the app hands your session over instead of making you start again. What happens: the app mints a one-time code and opens your browser. You see **Signing you in…** for a moment, then land on the page the app intended, already signed in. Three things to know: - The code is **single use** and expires in about 30 seconds. It cannot be shared or forwarded. - The hand-off only ever forwards you to a page inside Muin, so a tampered link cannot redirect you elsewhere. - If the code has already been used or expired, you are forwarded anyway and simply asked to sign in. Nothing is lost. If the browser sits on **Signing you in…**, go back to the app and tap the action again for a fresh code.
What does the results board on the Agents page show?
The **Agents** page opens with what your agents found and what it is worth: - **What your agents did for you**: issues caught and resolved in the last 30 days. Beside them, hours and money saved this month. The findings tiles open your Inbox; the savings tiles open Value. - **What your agents found**: one row per agent, with its open findings, high-priority findings, findings closed in the last 30 days and its last run. - **Value this month**: value found (at stake, not yet acted on) and value realized (confirmed), never added together. - **Findings by site, grant and severity**: every finding open now or raised in the last 30 days, with its owner, due date, whether it is open or closed, and the rule it cites. Filter by site, grant, severity and open or closed. Each finding opens the record it is about, and **Show your work** shows the facts and sources behind it. - **What your agents produced**: the latest reports, calendars and grant spend plans, each with the date it was produced. - **The next 12 months**: permits, inspections, filings, renewals, grant report dates and exemption days ahead. - **How much of this was checked**: how many statements were computed from your records, verified, not yet verified or removed. **Export CSV** and **Export PDF** download the findings that match your filters.
Why does a finding say its dollar figure cannot be computed?
A finding shows a dollar figure only when your own records support it, and it names the record it used. When a record it needs is missing, the finding says the dollar figure cannot be computed, says what is missing, and shows no number. Muin never estimates one. For the facilities audit: - **The property tax your exemption is worth** needs your property tax record. Add the assessed value, the levy rate per $1,000, the tax year and where the figures came from under **Settings → Facility Rentals → Property tax record**. - **The uncovered amount on an uninsured rental** needs the space's minimum coverage and, when a certificate is on file, its recorded per-occurrence limit. Add the missing record and run the agent again. On the facilities audit's top actions the same finding reads **No dollar figure yet** until then.
Can an agent edit my records on its own?
No. An agent can read your records and work out what should happen, but every change it proposes waits for a person to approve it. The same is true for Ask Muin, an agent step in a workflow, and Claude or any other AI app you connect. A change is anything that edits a record, or sends, posts, pays or files something outside Muin. Each one shows up as a preview: what would change, the record, the amount and who would hear about it. Nothing is written until someone clicks **Approve**. The one exception is a **standing rule** a person set in **Settings > Approvals**. It approves one kind of routine change in advance. Three limits always hold: - Only a person creates or changes a rule. No agent, Ask Muin conversation or connected AI app can. - Anything that pays or commits your organization to pay, such as a payment, a payout, a refund or an invoice approval, always needs a person to click approve. A rule cannot cover it. - Every change a rule approved is counted in your daily briefing the next day, with a link to **Inbox > Audit**, where **Undo** reverses each one that can be reversed. On an agent's **Settings** tab, **Always ask me when...** adds more questions, for example "the amount is over $1,000". These rules only add asks.
Can an agent pay a bill without me?
No. Anything that pays or commits your organization to pay, such as a payment, a payout, a refund or an invoice approval, always needs a person to click. An agent can prepare it and show you the preview, and it waits in your **Inbox** until you approve it. A standing rule cannot cover a payment or an invoice approval either. Muin refuses to save one: "Payments, payouts, refunds and invoice approvals always need a person to approve, because they pay or commit your organization to pay. A rule can't cover them." On your phone, a payment opens for review and is never approved straight from the notification.
What is a standing rule for agents?
A standing rule is your approval in advance for one kind of routine change, so you are not asked every time. Set it in **Settings > Approvals** with **Changes agents propose**. Choose which changes it approves and which agents it covers: a run a person started (this also covers Ask Muin and Claude), an unattended automation, or both. - Only a person can create or change a rule. Agents can't. - A rule can never approve a payment, a payout, a refund or an invoice approval. - Every change a rule approves shows up the next day in **Inbox > Audit** under **Approved by your rules**, ready to undo.
How do I undo something a standing rule approved?
Your daily briefing says "Yesterday your standing rules approved 3 changes" with a **Review** link. Or open **Inbox > Audit** and choose **Approved by your rules**. Click **Undo** on a row to put the record back the way it was. Undo is offered only by the person the rule acted for, or an admin, and it works once. Muin refuses, with the reason, when: - The change sent a message, changed another system or moved money. Those cannot be reversed. - Someone edited the record after the change. Muin never undoes over a later edit; fix it by hand.
Why is my agent's run waiting?
The agent proposed a change to your records, and a person decides every change before it is written. The run page reads "Waiting for … to approve the changes this run proposed." with an **Open in Inbox** link. Open **Inbox**, find the agent under **Waiting for your approval**, and click **Review**. Approve the changes you want (or **Approve all**) and reject the rest with a reason. Once every change is decided, the run finishes. On a run you started, the changes go to your Inbox. On a scheduled or automatic run, they go to the inbox of the agent's owner, the person who turned it on.
What happens if nobody answers an agent's change?
If the owner hasn't answered in 48 hours, the change goes to the agent's **backup approver**, and after another 48 hours to an admin. Someone who lacks the permission the change needs is skipped. Set the backup on the agent's **Alerts** tab under **Who approves this agent's changes**. The same block sets **Waiting changes expire after**: 3, 7 or 14 days, 7 by default. When a change expires, the run stops and nothing changes: "Expired: nobody approved within 7 days. Nothing was changed." The owner is told.
How do I check an agent change before I save it?
Open the agent's page and click the **Evaluate** tab. Click **Build from past runs** to turn up to 50 of its recent runs into checks. Leave **Reviewed runs first** on so runs someone approved or corrected come first. Open the set of checks, go to **Compare a change** and choose **What to try**: **Unsaved settings** (the model is a setting), or for a custom agent **A new prompt** or **A saved prompt version**. Click **Compare**. Each check runs twice, once as the agent works today and once with your change. Nothing is saved or sent. The results list **Regressions** first: checks that passed today and fail with your change. On a custom agent, **Compare before publishing** in the page header opens the same comparison with a new prompt.
Does comparing an agent use AI credits?
Yes. A comparison runs each included check twice, so it uses AI the same way the agent does. Before it starts, the **Compare a change** panel says "This will run N cases twice and use about X credits." The results show what it actually used. Comparisons are listed as **Quality checks** on the **Usage Breakdown** tab. A comparison runs only when you start one; saving or publishing an agent does not run one. Up to 50 cases per comparison.
What are practice checks when I build an agent?
Practice checks are ten made-up cases a new agent should get right before it runs for real: four everyday ones, three unusual ones and three it should refuse. On the review screen click **Write practice checks**, then **Run practice checks**. Each check shows **Passed** or **Failed** with the reason. Nothing is saved or changed while they run. If two or more fail, **Turn on** asks "Turn on anyway?" and records your answer. **Try the cheaper model** runs the same checks on Fast and says whether Fast passes every check your model passed. Practice checks are charged as building and testing credits.
Why does my agent say my corrections point to one change?
Three corrections on the agent's answers ("What should it have done?") said the same thing, so the agent's **Overview** offers one change to its instructions. Click **See the change** to read it, **Check it on past cases** to rerun the corrected runs with it, and **Apply** to save it as a new version. To undo it, open **Versions** and click **Roll back**. Each correction is also kept as a check on the **Evaluate** tab, in the **Corrections** set.
Why is a case marked Not reviewed?
Nobody approved or corrected the run the check came from, so its expected answer is what the agent produced at the time. A Not reviewed check tells you whether your change gives a different answer, not whether either answer is right. To make it count as the correct answer, click **Edit answer** on the check and say what a correct answer looks like. It then shows as **Edited here**. You can also untick **Included** to leave it out of comparisons.
What is the difference between Executive, Director and Manager Chief of Staff seats?
The tier decides how far down the reporting line a person's briefing reaches: - **Manager** ($50 per seat a month): the people who report directly to them. - **Director** ($75): two levels down, their team and their teams. - **Executive** ($100): everyone below them in the reporting line. Every tier has the same four sections: Team status, Approvals waiting, Risks and Value delivered. Buy and assign seats in **Settings > Platform > Assistants**.
How does a Chief of Staff briefing know who is on my team?
It follows each employee's manager in HR. An employee with no manager, a manager who has left, a reporting cycle, or a member with no employee record leaves people out of someone's briefing. Admins see these listed under **Reporting lines to fix** in **Settings > Platform > Assistants**, each linked to the record to fix.
Can a Chief of Staff seat see things its person cannot?
No. Every section of the briefing reads only what the person could already open, and every read is recorded in the audit log with the person it acted for. A section the person has no access to says so instead of showing as empty. A manager's briefing never includes a peer manager's team.
What happens when a Chief of Staff seat reaches its monthly AI credit cap?
The seat still gets its briefing, written from records only: no AI summary and no actions until the cap resets on the 1st or an admin raises it. The briefing says why. Set or clear a seat's cap with **Set cap** in **Settings > Platform > Assistants**.
Who can see AI activity?
Owners, admins and members whose role can view the audit trail. Give a member access by turning on view for **Audit trail** in their role. The page is at **Settings > AI activity** and only ever shows your own workspace.
How do I stop the AI fairness email?
An owner or admin goes to **Settings > AI activity > Fairness report** and turns off **Email the quarterly fairness report**. The report is still generated each quarter and stays in Reports; only the email stops. The other switch on that card, **Include our anonymous totals in the yearly public report**, controls the yearly public review, not the email.
Other FAQ Categories
Didn't find your answer?
Our support team is happy to help with any questions.
Contact Support