That’s also true. Miners like vitamin have a lot of money in bitcoin, if they tried to do this they’d hurt their balance a bit and also lose trust from their company.
Other companies like bitfury would have the equal impracticality of publishing the names and images of people working there, they’re probbly not too hard to track if they’ve stolen $5 million.
The current hashrate is at about 50TH/s. That's 50*10^12 hashes per second.
Now, lets assume they can try private keys in the same speed (which the never can, since you'd need to calculate the public key out of the private key, then hash it to get the address and then check if it cointains balance while mining simply is double-hashing).
They (all miner) would need ~ 1.15^63 seconds to check half of the keyspace. That's 1.34 * 10^58 days.
That's 1340000000000000000000000000000000000000000000000000000000000 days if ALL miner could search at the same speed as hashing works (in fact that's MUUUCH slower).
As you see, this would be completely wasted energy. It's simply not possible. The keyspace is waaaaay too big.
Individual miners are robably more towards the 100THs mark now and anyway, that’s just one miner not a whole pool that gets turned to mining addresses.
I don’t think the large bitcoin collider has done very well either in finding bitcoins (posted above by another user). It’d e header to write something to retrieve people’s data while they’re on the internet and get their private keys that way than probably mining them. Might be less time consuming too to go to every computer in the world and install your malware on it .