The Board did not decide. It waited until it no longer had to

The Board did not decide. It waited until it no longer had to.

Have you ever voted, on a Board, to ban something that was already running inside the company?

The table rarely decides on technology. It waits until it no longer has to decide.

July 2024. Fifteen to twenty people, partners and Board Members at the same table. The reference was Samsung: engineers uploaded internal code to ChatGPT in April 2023 and the company banned generative AI in May. Three people closed the ban, all three from multinationals. The tool was already running inside, on personal accounts, with no audit trail.

In the first half of 2025 I proposed the ugly way out: a single shared password, everything auditable. Bad security practice, and the only thing that exposed the real size of the usage.

What changed the Board came from outside: a corporate version, a closed environment, a team workspace. In 2024 it was a toy. Today it is a machine.

The Board did not change its mind. The risk was solved by someone else.

Blocked use does not become zero use, it becomes unrecorded use. Whoever comes in after the vendor has solved the risk arrives together with everyone else, and together means average. Two quarters stalled, almost a year to reopen, two years to use with force what was already in hand.

A table that only signs after the risk is gone is waiting for permission. And permission arrives for everyone on the same day.