To divide the responsibility for a bots actions between the person writing the code and the person running the code you must take intent into account. Anything can be used for good or for evil, what matters if what is the person behind the action intending to do and why do they want to do it. For example, if someone creates a bot to help people find research sources for a project and then someone uses it to find illegal things on the internet, it's the responsibility of the person who ran the program because they chose to abuse the program. However, to protect themselves the person writing the code could also add in preventatives to the code to prevent abusive usage. Yet, the responsibility in the example I provided still falls onto the person who chose to use the bot for 'evil' because they are the person actively seeking the 'evil.'

While Egoism seems dangerous on it's own, the basis of life relies on putting yourself first, after all you can't help anyone if you can't help yourself. An example of this is are the emergency oxygen masks on an airplane, to survive you must put on your mask first before helping anyone else, because if you don't then you could be off worse. Therefore, while in some scenarios Egoism is negatively impacting it's actually important to practice in moderation so that you don't neglect yourself - this practice is called Rational Egoism because it is the logical and morally correct pursuit.