A HBA1c level at 83 corresponds to an average bg of around 12.8 mmol/l.
With your own recorded readings to be just around 5-8 mmol/l, then it could indicate a couple of scenarios:
1. The bg reader does not work (highly unlikely, especially if you test/compare with your backup meter).
2. The tests you do take, are maybe just done 'when knowing you are in the safe zone' with good results. It is certainly quite suspicious if you only have readings in the 5-8 mmol/l range, as that is almost impossible with T1 except for a seasoned low carber and/or sitting still without moving and just eating the same day in day out.
3. The bg reader results/logfile is being tampered with. Either dilution of blood sample and/or correction in the measurements database.
Highest chance for number two. If you want to experiment, measure bg before meal, but also like 1-2 hours after. Personally I always measured before, to decide insulin volume for the meal, but rarely after, despite I knew I might go through some spikes. I only had bg test sticks for in average 4.5 tests per day, so found it most important to measure before meals and not after. And then again measuring just before going to bed. So meter analytics indicated lower average bg over 3 months period compared to the HBA1c indicated for same period. Though must admit, not as big difference as you indicate in your posting.