Here's a graph from johnstonsarchive.net showing energy released by earthquakes and, however reliable this graph is, it shows huge differences in earthquake activity. About how big percent is the energy of quakes caused by humans?
You have an interesting graph there. In response to your question, I would say that if we actually use the Mw (moment magnitude) measure of seismic activity (which is related to energy release via slip on a fault), we would see that most of the Earthquakes 'caused' by humans (and note that there are still only a few examples of this, (such as the 5.7 Prague, Oklahoma Earthquake), are limited to intraplate earthquakes which only exceed 4 in rare instances such as the above.
The exact 'percent' of energy release due to induced (human-caused) seismicity would probably not even approach single digits, but possibly reside in values approaching zero, when you consider that the vast majority of earthquakes take place at plate boundaries and outnumber induced earthquakes by a large amount.