Your own words
AI does not write, complete or rewrite a colleague’s message.
Team occasions
People write.
AI checks.
AI content moderation for workplace group cards. tbf work checks notes for threats, abuse, spam and private details, with a person able to review flagged notes.
Get started for FREE! then £4.99 per card
AI checks. Human review. Your own words. ↓AI content moderation for group cards
What the card check does.
The check is about a note’s content. It does not compose messages, decide what colleagues should feel or rate an employee’s performance.
- 01
Looks for harmful content
The checks include threats, hostile abuse, mockery, sexual content, spam, unrelated material, attacks on somebody else, private details and attempts to steer the check.
- 02
Keeps flagged notes off the card
A flagged note is held separately from the messages the recipient reads. It can then be reviewed by the person allowed to manage that card.
- 03
Lets a person correct it
Company managers and admins can restore or delete a note from the spam folder before writing closes. On a single card bought without an account, the creator reviews set aside notes.
- 04
Keeps writing human
The check never writes, completes or rewrites the message. Restoring a flagged note puts back the contributor’s original words.
- 05
Has an explicit fallback
If the AI service cannot answer, the note is accepted. A check can miss harmful content or flag an ordinary message wrongly; it does not guarantee a safe card.
The organiser’s list
Before sharing the link.
Share the writing link with the intended contributors. Check the review area before writing closes and keep sensitive details out of the card yourself.
- Choose the intended writing group.
- Know who can review this card’s flagged notes.
- Read the privacy notice before sharing personal information.
- Review flags while the card still takes writing.
- Do not use the card check for employment decisions.
AI safety checks
People write.
AI checks.
The check looks for threats, abuse, sexual content, spam and exposed private details. Flagged notes are kept off the card for a person to review.
How the checks work ↗A person can correct it
Company cards have a spam folder for managers and admins. On a card bought without an account, the creator reviews set aside notes through their manage link.
The limits are clear
AI can miss a note or flag one wrongly. If the check is unavailable, writing still works.
Check the data policy
Notes are sent to TypeSafe for the check. Names and card access codes are left out of that request.
Read the privacy notice ↗From an occasion to a card
Give them
everyone’s words.
- 01
Create the card
Pick a cover, name the recipient and choose when it opens. You can buy one card without a company account.
- 02
Share the writing link
Colleagues contribute in a browser. Each person writes and signs their own message.
- 03
They open it on the day
The messages arrive together in one card. The recipient reads it on their phone.
For repeated birthdays, work anniversaries, welcomes and leaving days, use a company account with Slack and BambooHR.
See the plans ↗Questions
Which provider checks the notes?
TypeSafe’s Jev checks the note and its occasion. The request leaves out contributor and recipient names and the codes that give access to the card.
Who can see flagged notes?
Company managers and admins see flagged notes on company cards. The creator sees set aside notes on a card bought without an account through their private manage link.
Does the AI guarantee a safe card?
No. It can miss content or flag a note wrongly. Notes also go through if the check is unavailable, and a person can correct a flag before writing closes.
See what the recipient reads.
Get started for FREE! then £4.99 per card
Checked 30 September 2026.