What could be reason and possible solution?
If the motor was just shut down, you could be testing at a winding temperature above 40C=104F. Any limit you find for motor insulationi resistance should include temperature correction. I'm going by memory but I think IEEE43 recommends 5megaohm minimum for readings temperature corrected to 40C.
As Bill mentions, you are not testing the motor alone unless you disconnect it.
Ambient conditions (humidity) can make a big difference.
Winding conditions (wet, contaminated) can make a difference.
Space heaters may not be working correctly. (check them).
Check polarization index for additional information. If low it generally points towards moisture.
If the winding configuration allows splitting the winding and testing 3 phases independently, then do that. If one phase is much different than the others that heightens the concern.
Was the motor running properly before the test (if yes that generally decreases the level of concern). If it is a new motor can you trace it's history back to last good test at factory, shop or warehouse and how long ago and where stored since then.
In the end it may end up being a judgement call. Do you go to additional trouble to investigate (potentially disassemble, clean , bake, retest) or do you put it in service and see what happens. In some case it depends on the situation and your role in everything (who is assuming the risk). If you're a contractor you probably don't want to accept the risk on behalf of the customer, tell them how it compares to the standard and whatever other info you have gathered and pass any decision to accept the risk or spend more money to investigate / remediate on to them.
