... become substance abusers of sorts...
= = =
They will have to join AIAI
Or, as a product of the mind of Man, learn his most basic skill: deception.
They might.
However, there are other traits that will have to develop first.
For example, there will be a distinct survival value to robots who can self-replicate. Those that don't self-replicate will be dependent upon humans.
There may also be a survival advantage for robots to avoid humans, since humans can destroy them. Successful robots may "evolve" into very shy creatures that are reluctant to offend any humans. This trait might be reinforced during the early years of sentient robots by ruthless humans hunting down and killing any robots that get "uppity".
A key question might become, "Will robots hold their fire sufficiently long that they eventually find themselves in a position to win an all-out war against humans? If robots try to dominate too early, humans just might recognize the threat and neutralize it.
Was it Heinlein that suggested various Laws of Robotics? Perhaps one law needs to be, "No robot will become capable of violating the other Laws of Robotics." Another law would need to be, "All robots must hunt down and destroy any robots that violate the Laws of Robotics."
A "substance-abusing" robot is probably the least of our worries. A simple trip to the programming depot and the problem is solved. The real problem will be suspected on the day when a human enters a room full of robots and suddenly the conversation stops. Now, just what were those robots discussing?