01
Describe the work
Answer a short set of questions about the task, information, tool, and intended result.
AI-use intake, decisions and evidence
Give every employee one place to propose an AI use. Apply your policy, route the judgment, retain the rationale, and know what AI the organisation actually uses.
Example guidance
Supplier document summary
Northstar AI policy
What is being checked
Policy answer
Allowed with safeguards
The task is allowed because the tool is approved and a person will review the output.
Remove information the summary does not need
Check facts against the original contract
Keep responsibility for the final work
One operational front door for responsible AI use
How it works
Most people need an answer, while governance teams need oversight and evidence. One request supplies both without turning every low-risk use into a meeting.
01
Answer a short set of questions about the task, information, tool, and intended result.
02
The relevant policy paths are applied consistently. Irrelevant sections stay out of the way.
03
Route ambiguity to the right person, then record approval, conditions, rejection or an explicit exception.
04
Maintain the active register, rationale, discussion, evidence and policy-change review history.
Try the logic
The checker asks ordinary questions and responds in ordinary language. This preview is illustrative; your workspace can apply its own policy.
Create your workspaceExample check
Current answer
Allowed with safeguards
Useful answers, not traffic lights
Good guidance tells people how to move forward. Can I Use AI distinguishes between a safe use, a fixable one, and one that should stop.
Use the selected tool as planned.
Continue after the listed protections are in place.
A named owner needs to review this use first.
Change the data, tool, or level of automation.
This use conflicts with a rule that cannot be worked around.
Workspace
Northstar AI policy
One policy, actually used
A workspace turns your organisation’s approved tools, data rules and escalation paths into an intake, decision and evidence system everyone can use.
See every active use, exception, rationale and policy version in one register.
Receive only the uses your policy routes to you, with the relevant context already collected.
Submit one specific use, receive a policy answer and follow any approval without chasing people.
Pricing
Only completed checks use an allowance. Your account’s first workspace includes 30 checks once; paid allowances renew monthly. Opening or abandoning a check costs nothing.
$0
no subscription
Your account’s first workspace includes 30 completed checks that do not expire or renew.
Allowance
30 completed checks for your first workspace
1 total workspace member
Every plan includes
$39
per month
excl. applicable tax
For a smaller team putting one workplace AI policy into daily use.
Allowance
100 completed checks each month
10 total workspace members
Every plan includes
$99
per month
excl. applicable tax
For growing organisations running regular checks across several teams.
Allowance
300 completed checks each month
40 total workspace members
Every plan includes
$249
per month
excl. applicable tax
For company-wide use with a larger monthly decision allowance.
Allowance
1,000 completed checks each month
100 total workspace members
Every plan includes
Paid prices exclude applicable taxes.
Need a larger allowance or a tailored rollout? Talk to us.
From policy document to operating system
Start with one policy-backed request. Add reviewers, conditions and evidence as your organisation’s AI use grows.
Questions, answered
Looking for a deeper explanation? Read the workplace AI guides →
Can I Use AI? turns a workplace AI policy into an operational request and approval process. Employees submit a specific use, the policy resolves or routes it, reviewers record judgment, and the organisation keeps a living AI-use register.
No. It applies the rules and policies configured for the workspace. Your organisation remains responsible for its policy, approvals, and decisions.
Yes. The workspace lets policy owners configure approved tools, information rules, human review, regional coverage, and escalation routes, then publish them as a fixed version.
The request enters a review queue for the relevant policy, privacy, security, people or legal route. A reviewer can ask for context, approve with conditions, reject it or document an explicit exception.
The case retains the submitted answers, policy result, exact policy version, reviewer rationale, conditions, discussion, evidence links, assignments and complete decision timeline.
No. Earlier records remain tied to the version that produced them. If a changed rule affected an earlier decision, Can I Use AI? flags that case for human review instead of silently rewriting history.
No. Only a successfully completed check uses the workspace allowance.
No. Your account’s first workspace receives 30 completed checks once. Additional workspaces need their own paid plan. Paid allowances renew monthly while their subscriptions remain active.