Hi, I understand the 26V with 47k series and 11k shunt, which should give you about 5V at the pin. And 47k series with 47k shunt will give half of what the applied voltage is. However, the MCU pins have diode clamps on each pin, so the input voltage (at the pin) will be no more than a diode drop above VDD. If your VDD is 5.0V, the max pin voltage would be about 5.7V, assuming you limit the current.
I tried a simple experiment on multiple port B pins and have not seen any problems. With VDD at 5.0V, the source voltage at 26V, and 47k series with 47k shunt, I measure 5.66V at the pins. The pins are still functional.
The AW parts can allow a certain amount of injected current - current that flows into the pin when its voltage is higher than VDD. This current is on the order of 1 - 2mA. In my experiments the injected current was around 0.3mA. That's not enough to destroy the pin.
What do you mean when you say the MCU is broken?