We won't know until the telemetry and video is analyzed but it appears that the booster stage did not get the expected engine burn pattern after staging. This would cause the trajectory to go outside the flight safety envelope which would initiate flight termination.
The second stage data requires more analysis. The only thing I could see was premature thrust termination.
Compared to the Atlas program that I worked on, these flights have been very successful considering the learning curve is usually based on failure analysis. The Atlas had so many initial failures that when one finally worked the launch crew renamed the missile the "AT LAST".
LOL! And thanks for that anecdote :-)