Evidence findings
Gaps, conflicts, or limits found in the information and commitments you supply.
Communication assurance for HR, communications, and change teams
For the difficult moments you're trusted to get right.
Find the risk before
the message goes out.
For communications professionals, HR and people teams, change teams, managers, and leaders, AIClearline is human-led communication assurance and decision support, not another writing tool. It surfaces material risks and cross-channel inconsistencies while people retain authority over every decision.
Build and review your first program free for seven days. No card required. A subscription is required to export or share work or continue with another program.
Collective redundancy
“We have decided to remove 40 roles. Consultation begins Monday.”
The word “decided” makes consultation sound performative. It can become evidence that the outcome was fixed before alternatives were heard, and an all-staff announcement may be inappropriate before affected people are notified.
Corrected in review
“We are proposing changes that could affect up to 40 roles. No final decisions will be made until consultation is complete and alternatives have been considered.”
Illustrative UK collective redundancy scenario. AIClearline flags communication and sequencing risks; current UK employment-law and employee-relations review is still required.
AIClearline in 15 seconds
See the judgment,
not just the claim.
One sentence. One unsupported assurance. One human decision.
You provide
01“No roles are affected by this change.”
AIClearline flags
The Approved Position only confirms that no role changes are planned in this phase. “No roles are affected” creates a broader assurance than the evidence supports.
Why it matters
03If the scope changes later, employees may reasonably interpret the original statement as a broken commitment.
Human decision
04KeepReviseAdd evidence
AIClearline challenges the Approved Position
Not a black box. See the methodologyThe method before the message
It establishes and challenges the Approved Position before drafting
The method brings together a clear communication workflow, relevant audience perspectives, and integrity checks, while keeping human decisions explicit.
- 01
Establish
- 02
Challenge
- 03
Decode
- 04
Align
- 05
Prepare
- 06
Decide
- 07
Learn
Three kinds of finding, clearly labeled
Methodology findings
Issues surfaced by the framework's checks, with reasoning for a person to review.
Audience hypotheses
Plausible ways relevant audiences might read the position, never predictions of how they will react.
The framework supports human judgment. It does not promise to catch every risk, predict outcomes, provide legal advice, or guarantee that a control will work.
Illustrative program view
From situation to decision-ready internal communications
You describe what's happening. AIClearline reads it and drafts an Approved Position. You review, approve, and lock it before the assets the situation calls for are generated. Each one is yours to review, edit, and approve.
1
AIClearline reads your situation and names its risks
Situation read: EMEA Restructure
Emotional temperature
HIGH
Trust level
LOW
Manager dependency
REQUIRED
Risk flags
Prior commitment at risk. Your last town hall included a “no compulsory redundancies” statement. Employees may compare this announcement with that promise, especially if you do not acknowledge it first.
Unaffected employees need their own message. Without explicit communication, staff in unaffected regions may wonder whether they are next. They may read silence as uncertainty.
What is actually happening
This is a trust-critical announcement in a fragile environment. The proposal could affect 20% of EMEA roles, so employees may focus on that scale regardless of the intended framing. The communication challenge is to explain the proposal while addressing who may be affected, who may not be, and what happens next.
Approved Position locked · Tested from the audience perspectives that matter to this situation
2
One Approved Position. Six sequenced drafts, yours to review.
Communications: EMEA Restructure6 sequenced drafts · not final
1
Draft · awaiting review
Union / representative formal notification, where required
2
Draft · awaiting review
Manager briefing before at-risk conversations
3
Draft · awaiting review
At-risk consultation letter, delivered in a 1:1
4
Draft · awaiting review
All-staff email after individual notifications
5
Draft · awaiting review
Survivor communication
6
Draft · awaiting review
FAQ
Inside the drafts
We previously said “no compulsory redundancies”. This proposal will be considered through consultation before any final decisions are made.
Draft all-staff email · prior commitment explicitly reconciled
We are not proposing changes to roles in [defined scope] at this stage. Consultation is under way, no final decisions have been made, and confirmed updates will follow through the consultation process.
Draft survivor communication · scope remains conditional during consultation
Draft consistency checks completed. Not a readiness decision or professional review. Current evidence required: employment law, employee relations, and dignity.
Illustrative asset set
Start your first program
Same announcement, different possible readings. The Approved Position is stress-tested from relevant audience perspectives before drafting, with dedicated assets where the situation calls for them.
Build and review your first program free for seven days. No card required. A subscription is required to export or share work or continue with another program.
You approve the Approved Position
No communication asset is drafted until you approve the Approved Position, address findings from relevant audience perspectives by revising and re-testing or explicitly accepting them, and pass the readiness check. The plan adapts that Approved Position across the audiences and channels the situation requires, and shows which human actions must happen before wider communication.
You see plausible perspectives
Before drafting, the Approved Position is stress-tested against relevant audience perspectives. Those are plausible readings, not predictions. Each asset is then checked against the Approved Position and its claims; deeper executive reads are available when you request them.
You decide when it goes out
AIClearline does not autonomously send to employees or publish through customer channels. You decide what is approved, what is exported as working material, and when any final communication reaches its audience.
AI advises
Situation read
You describe what's happening. AIClearline reads what you wrote, and what's between the lines. The tension underneath the announcement, the promise employees will remember, the question you haven't answered yet, the audience you haven't thought about. You review the diagnosis before approving the Approved Position or drafting communication assets.
You lock it
Approved Position
After you describe the situation, AIClearline drafts an Approved Position from the information you provide and identifies what is confirmed, what it means, what is not changing, and what remains unknown.
You review it, approve it, and lock it. Every communication that follows is built from that Approved Position: not the same words repeated, but the Approved Position shaped for how each audience will hear it. The manager who has to answer questions needs different words from the employee wondering if they're next.
You see every verdict
Stress test
Before communication assets are drafted, AIClearline tests the Approved Position from the audience perspectives that matter to the situation. These can include a skeptical reader, an employee wondering whether they are affected, a manager preparing for questions, and someone who feels disconnected from the change.
Same facts. Different possible readings. You see the perspectives and decide whether to revise and re-test the Approved Position or explicitly accept the findings before communications are drafted.
These are plausible perspectives to consider, not predictions of how employees will react.
You approve the checked drafts
The assets each situation needs
Manager briefings, all-staff emails, FAQ pages, listening guides, employee share messages, and digital screen copy can be generated across the product's channels, with required scenario assets added where applicable. Each begins as a draft for your review. You can approve checked, clean drafts individually or use batch approval for eligible assets.
Plus the tools that keep the program honest: commitment register, program report, and consistency audit.
AI flags it first
Consistency enforcement
AIClearline checks each communication against the Approved Position and consistency locks you approved, then flags consistency, out-of-context, and claim issues for review. It helps keep drafts aligned; approval remains yours.
You decide what's next
Measurement layer
Three questions, once a week. How equipped did managers feel? How many questions escalated to HR or leadership? Were employee questions fearful, skeptical, practical, engaged, or were there few or no questions?
When available, the AI interprets those customer-recorded signals together: what people may have heard, what is still missing, and what may need to change. You get the interpretation. The decision is yours.
Signals are recorded by the customer. When available, AIClearline interprets them for consideration. They do not, by themselves, prove that a message landed or changed outcomes.
15 configured output languages
Write in the language your people communicate in
Output-language configuration supports 15 languages. Generated text and translated interface content still need review by a qualified speaker; semantic quality and every translated interaction are not verified equally across all 15 languages.
If your situation description doesn't say what languages your workforce speaks, the AI will ask, because people can't hear an announcement written in someone else's language.
How it works
01
Just say what's happening
Describe it the way you'd tell a colleague. What's going on. Who's affected. What you're worried about. No finished brief, template, or communications background required. The guided intake helps you build the brief as you go, for the people carrying the announcement and everything riding on how it's received.
02
You approve the Approved Position, then test how it may be read
AIClearline drafts the Approved Position from your guided brief. You review, approve, and lock it. Before communication assets are drafted, the Approved Position is stress-tested from the audience perspectives that matter to the situation. You decide whether to revise and re-test it or explicitly accept the findings.
These are plausible perspectives to consider, not predictions of how employees will react.
03
You review, approve, and decide what goes out
Communication assets begin as drafts built from the Approved Position you approved and shaped for their audiences and channels. Checks flag consistency, out-of-context, and claim issues. Approve checked, clean drafts individually or use batch approval for eligible assets, then decide what to export and deliver through your own channels.
04
You record what happened. You learn what to change.
After the communication goes out, record what you observed: whether managers felt ready, which questions escalated, and what people were asking. When available, AIClearline interprets those customer-recorded signals and carries saved learning into your next program. You bring the evidence. The advisor helps you decide what to do next.
You'll make decisions along the way. That's the point. You make every call.
After the words are approved
Move approved content into your tools, then record the signals
Approval isn't the finish line. Add-ins put approved content into a Word document or Outlook draft. After you deliver it through your own channels, record what you observe. When available, AIClearline interprets those signals for consideration.
Word and Outlook add-ins, host verification pending
Add-ins provide paths to browse approved communications and insert their content into Word documents or Outlook drafts, with headings and emphasis preserved. Microsoft-hosted Word and Outlook insertion is not verified for launch; confirm host compatibility before relying on it.
- Only approved assets are available to insert
- You remain the sender
Record the signals after delivery
After you deliver the communication through your own channels, record three weekly inputs: manager readiness, escalation count, and question character. When available, AIClearline interprets patterns and offers considerations or possible next steps, not proof of impact.
Manager readinessEscalation countQuestion character
Read · Approved Position tested · drafts checked · approved · exported · signals recorded: one governed record.
Why trust the read
It keeps the Approved Position accountable
Recommendations and drafts draw on the information you supply, the Approved Position you approve, and relevant audience context. The record preserves the Approved Position, checks, and decisions so you can review alignment as communications develop.
It flags, it never flatters
The Approved Position challenge and asset checks are designed to surface weaknesses, assumptions, omissions, and communication risks for review. Current audience-perspective findings remain visible, including when a human explicitly accepts them. You decide how to respond and make the final approval decisions.
It shows the basis for every flag
Asset checks identify relevant text or evidence and explain the concern. You can check that basis against your own situation before you act.
It learns from what you record
After you deliver the communication, record manager readiness, escalation count, and question character. When available, AIClearline interprets patterns for consideration and carries saved learning forward. These signals are not proof of impact.
Reasoning you can inspect before delivery and signals you track afterward give you evidence to review as the situation develops.
Why not use only a general-purpose AI chatbot?
Because the draft was never the hard part
A general-purpose chatbot can help draft from the prompt and context you provide. AIClearline adds a governed workflow for the situation, Approved Position, relevant audience perspectives, draft checks, approvals, and customer-recorded signals.
A general-purpose AI chatbot
- Can draft from the prompt and context you provide
- May not provide this workflow's governed situation, review, and approval record
- Does not inherently apply AIClearline's commitment and asset-check controls
- Leaves you to establish your own review, approval, and delivery process
AIClearline
- Reads the situation you describe and flags risks or missing conditions for your review
- Checks drafts against the Approved Position and consistency locks, including commitments captured there
- Stress-tests the Approved Position against relevant audience perspectives
- Generates situation-required drafts, checks each against the Approved Position and locks, and, when available, interprets customer-recorded signals
AIClearline diagnoses the situation and drafts the plan and communications it calls for. It checks each draft against the Approved Position and locks, supports manager-facing drafts and actions, and, when available, interprets customer-recorded signals for consideration. You make every call.
AIClearline does not make your decisions for you. It structures the choices around the situation, Approved Position, audience perspectives, and drafts. You approve the Approved Position, review drafts and perspective findings, decide whether to revise or accept findings, and decide what to approve, export, and deliver through your own channels. You make every call.
Security & confidentiality
Built for the most sensitive internal communications you'll ever send
Restructures, departures, pay changes: the situations you bring to AIClearline are confidential by nature. Here's how they're treated.
Never used to train AI
Your content is processed by Anthropic's Claude. Under Anthropic's commercial terms it is never used to train their models, and we never use it to train any model of our own.
Confidential content stays in the app
The work you do here is never copied into our email or marketing tools. When we email you, the message says only that something needs your attention, with a link back to the app. An email of ours reveals nothing about what you are planning.
Nothing sent without you
AIClearline does not autonomously send to employees or publish through customer channels. You choose exports or insertion into external host documents and control distribution.
Encrypted and limited to your organization
Your data is encrypted while it travels between your browser and AIClearline. People must sign in, and the app limits each customer to their own organization’s information. This separation is enforced in the software rather than through a dedicated physical system. Passwords are securely hashed and never stored as readable text.
What account deletion removes
Deleting your account removes eligible product data from AIClearline's database. We initiate cleanup with our payment and email providers, but provider-side work may remain pending, and records they are legally required to keep may remain.
Analytics only if you opt in
We run no third-party tracking unless you accept it. Google Analytics loads only with your consent. Core product use does not require optional analytics.
Accountability
AI helps. A named person is accountable.
When AI helps shape a sensitive communication, the questions that matter later are simple: what was approved, who approved it, and when. AIClearline is built to keep that answer on the record.
You hold the pen
AIClearline drafts, stress-tests, and flags risk, but it never sends or publishes anything. A named person reviews, edits, and approves every communication before you send it.
Approvals are on the record
Every sign-off is captured in your program's history: who approved, what they approved, and when. Accountability here isn't a policy statement. It's a log.
A record you can produce
If you're ever asked how a communication was created (by leadership, HR, or a regulator), you can export the record: the Approved Position, the risk checks, and the named approvals.
This describes how AIClearline works. It is not legal advice, and no software alone can make an organization compliant with any law.
The complete workflow, at the scale you need
Every plan includes the core judgment and communication workflow. Plans differ in active-program capacity, support, and organization workspace features.
Essential
$99/month
No monthly program limit. One active program at a time: finish one, start the next · 1 user
Professional
$299/month
Multiple active programs at once, with no concurrency limit · 1 user
For organizations
Enterprise
A governed system for the judgment behind high-stakes organizational communication.
A governed organizational workspace for five named practitioners, including the owner, with organization-owned records and shared institutional memory.
Separate accounts are not equivalent: they fragment knowledge and cannot preserve one accountable record. AIClearline complements distribution platforms, rather than replacing them.
$12,000/ organization / year
Can I run 30 programs in a month on Essential?
Yes. Essential has no monthly program limit. Run as many as you need, one active program at a time. Finish one, then start the next. Choose Professional only if you need multiple active programs at once.
Every plan brings together three layers of assurance: a structured communication methodology, challenge from the audience perspectives that matter to the situation, and integrity checks that keep the Approved Position and every draft aligned. You approve the decisions and control what goes out.
Build and review your first program free for seven days. No card required. A subscription is required to export or share work or continue with another program. Refund policy
Does AIClearline predict reactions?
No. Audience-perspective findings are plausible readings, not predictions of how people will react. That is more defensible: it makes interpretations and their basis reviewable without claiming to know what people will do.
Is AIClearline a chatbot?
It asks questions first to understand the situation. You approve the Approved Position, decide whether to revise or accept audience findings, and approve each draft. You make every call.
Know someone who sends hard messages?
Passing this page to a colleague in comms or HR is the finest compliment you can pay us.