For IT services and consultancies
Answer the disclosure clause with evidence.
Your MSA now asks you to disclose how generative AI contributed to a deliverable. Most suppliers will answer that with a policy statement, because a timesheet only records who booked the hours. WhoWorked gives you the actual record, ready to hand to a client, a prime, or an auditor.
- Runs alongside your PSA
- Client-ready disclosure exports
- SSO and audit exports on Enterprise

What you are being asked to sign
Disclosure clauses are easy to sign and hard to satisfy without a measurement.
Enterprise MSAs increasingly require suppliers to disclose how generative AI contributed to a deliverable.
That measurement has to exist before the question is asked, and it has to survive a client, prime, or internal review.
Contract excerpt
“Supplier shall disclose the extent to which generative AI tools were used in the production of any deliverable.”
The shape of clause now appearing in enterprise MSAs
Where the record breaks
Three gaps between what you invoice and what you can evidence.
The MSA started asking
Client contracts now carry AI-use clauses. Signing one means attesting to how AI was used, which most firms cannot evidence today.
Who did the work is still an assumption
A T&M timesheet records the person the hours were booked under. It does not record whether an agent produced the output behind that booking.
Subcontractors widen the blind spot
Bench and partner work already arrives with the least provenance. Add agents to that chain and the gap between invoice and evidence gets wider.
What the record gives delivery leaders
Make provenance exportable instead of reconstructing it from memory.
When a client asks what AI was involved in an engagement, the answer should come out of the delivery record, not out of asking delivery leads to remember the week. Every entry carries the tool, the model, the tokens and the cost behind it.
- Keep human hours and agent sessions on one entry, with model and cost attached.
- Trace any deliverable back through people, agents and subcontractors.
- Export attribution for a client, a prime, or an internal audit.
- Answer disclosure clauses from evidence instead of recollection.
Attribution export
Northstar portal migration · PDF
14 AI sessions · $25.20 model spend · ready for client or prime review
Bring one engagement into the record first.
Start with a single project, connect the agents already running on it, and leave everything else where it is. If a security or procurement review has to come first, tell us your timeline and we will route it.
Turn attestation into a workflow.
01
One record, not two systems
Human hours and agent contributions land on the same entry, with tokens, cost and model attached. No reconciliation between a timesheet and a separate AI usage report.
02
Provenance you can hand over
Trace any deliverable back through the people and agents that produced it, then export it for a client, a prime, or an internal review.
03
Attest from evidence
When procurement asks how AI was used, the answer comes from the delivery record instead of a last-minute spreadsheet: which tools, how many sessions, what they cost, on which deliverables.
FAQ
Start counting all the work.
30-day free trial. Bring the time history you already have.