Unable to do simple addition of columns to table in Alembic migration

I'm using cockroach cloud and I'm attempting to do a simple migration where I add columns using alembic but I'm seeing this error:

sqlalchemy.exc.DBAPIError: (sqlalchemy.dialects.postgresql.asyncpg.Error) : this schema change is disallowed because table "my_table" is locked and this operation cannot automatically unlock the table
DETAIL: To unlock the table, execute `ALTER TABLE my_table SET (schema_locked = false);`
After the schema change completes, we recommend setting it back to true with `ALTER TABLE my_table SET (schema_locked = true);`

The cluster settings already have schema.auto_unlock set to true

all this does is add a few columns with null default

I opened up cockroach cloud, went to the SQL Console and typed in the command to unlock it explicitly and got this error:

ERROR: executing stmt 1: run-query-via-api: "schema_locked" can only be set/reset on its own without other parameters in a single-statement implicit transaction.
SQLSTATE: 22023

I am guessing the cloud cluster upgraded between now and the last time I attempted to modify a schema on a table because this historically has never been an issue. I don't reproduce this issue running the migration against a test cockroachdb instance which is v24.1.2

Jira issue: CRDB-65863

No repository file or test entry point is named. Start by reproducing the Alembic migration on Cockroach Cloud and on the reported v24.1.2 test instance, then compare the schema_locked behavior with CRDB-65863 and the SQL Console error. Done means the nullable-column migration has a documented or verified resolution without the current table-lock failure.

Written by the indexing model from the issue text.