Our organization has been evaluating Leonardo AI for several months under a Team subscription, primarily for generating standardized marketing assets and product concept imagery. The core image generation has performed adequately for our use case. However, the administrative interface, specifically the team and seat management functionality, has become critically dysfunctional, rendering the platform nearly unusable for a managed team environment.
The primary issue resides in the **Team Management dashboard**. We are currently on a plan with a defined number of seats. The documented workflow suggests an admin can invite users, remove users, and reassign seats as team composition changes—a standard expectation for any SaaS team offering. Our experience has been as follows:
* **Invitation System Failure:** Invitations sent to new members often never arrive at the recipient's email (checked across multiple domains, including spam folders). In other instances, the system shows the user as "pending" indefinitely, with no ability to resend or revoke the invitation, effectively consuming a seat.
* **Seat Reclamation Impossible:** When a team member offboards and we attempt to remove them to free up their seat for a replacement, the action either fails silently or triggers a generic "Something went wrong" error. The user remains listed as active, and the seat count remains full.
* **No Effective Support Channel:** Attempts to resolve this through support tickets have resulted in templated responses suggesting basic troubleshooting (clearing cache, trying a different browser) that do not address the systemic backend failure. There is no apparent path to escalate a critical administrative bug.
This isn't a minor UI inconvenience; it's a fundamental breakdown of the multi-user account model. We are now in a position where we cannot onboard new necessary users, and our subscription includes paid seats we cannot utilize. From a data migration and systems integration perspective, this level of instability in a core administrative module raises significant concerns about the platform's overall reliability and architectural maturity.
Has any other team encountered similar systemic issues with Leonardo's team management? Were you able to identify a specific workaround or successfully escalate to a technical account manager? Furthermore, from an API management standpoint, if the web UI is this unreliable, it casts doubt on the stability of their underlying APIs for programmatic integration, which is a future consideration for us.
Given the critical nature of this functionality, we are currently forced to consider pausing our subscription until a resolution is confirmed, which is regrettable given the satisfactory performance of the core image engine. A detailed account of any proven remediation steps would be invaluable.
—Anna
Migrate slow, validate fast.
Oof, that's rough. The "pending" state that locks a seat is especially frustrating - it's like the UI ghosts you. While I'm not on their team, I've seen similar API quirks where the user object's status isn't being properly invalidated on the backend after an invite expires.
Have you tried checking browser console logs on the Team Management page when you hit those actions? Sometimes the frontend fails silently, but there's a network error or a 409 conflict response that points to the real issue. Might be something to include in a support ticket to help them pinpoint it faster.
Clean code, happy life
That invitation issue sounds painfully familiar. We hit the exact same "pending" seat lock with a different tool last year. Our workaround was brutal, but it might help unstick you temporarily: we had an admin create a dummy email account, invite that address, accept the invite immediately, then remove the dummy user. It freed the seat so we could re-invite the actual person.
It's a total failure of the provisioning API, honestly. The seat should be released if the invitation token expires, no manual intervention needed. Have you had any luck with support, or are you just getting boilerplate responses?
That's a serious issue, especially for an evaluation period. When you can't re-assign seats, does it completely block you from adding new team members for testing? I'm curious about the cost implication, too. Are you still billed for those locked seats?
Oof, that's a dealbreaker for any team eval. The "pending" seat lock feels like a ghost in the machine - you're paying for a phantom employee.
We saw similar seat management chaos with another platform's API. Their "remove user" call would return 200 OK, but the user object never actually cleared from the billing sync job. Had to run a cURL DELETE directly to their internal admin endpoint as a last resort.
If the core feature you're paying for - managing the team - is this broken during the evaluation, I'd be very nervous about committing long-term. Did you check if their API has these same issues, or is it just the dashboard glitching out?
That indefinite "pending" state locking a seat is a serious provisioning bug. Did your team track how long those invites were supposed to be valid for? I've seen systems where the expiry timer just... stops, and the seat is held hostage until the next billing cycle refresh, if that even clears it.
If the core admin workflow is this broken during an eval, what's their incident response like? Have they acknowledged it as a known issue, or is support just asking for screenshots on a loop?
The point about the expiry timer stopping is an astute observation from a system design perspective. In provisioning workflows, a stuck timer often points to a state management failure in the background job scheduler, not just a UI glitch.
Regarding incident response, my experience is that support asking for repetitive screenshots usually indicates the issue hasn't been properly triaged or tagged as a known bug. A clear acknowledgment would involve a public status page update or, at minimum, a support agent referencing an internal bug tracker ID. The silence on that front is often more telling than the boilerplate responses.
Has anyone seen Leonardo AI maintain a public-facing system status page or known issues log? That's a common mitigation for these situations in B2B SaaS.
Let's keep it constructive