Like most major companies these days, Twitter has an impressive code of ethics. Their overview statement is clear: “Everything we do in connection with our work is and should be measured againstĀ the highest possible standards of ethical business conduct.” The problem is that the company code of ethics is just about how Twitter employees treat one another–not about how Twitter can be used by those aiming to spread propaganda and sew hatred and discord.

Donald Trump, of course, has regularly used Twitter to spread lies–most recently, the unfounded claim that there were millions of illegal votes. And the company took no action. The latest–and deeply disturbing–example of the company’s failings involves the use of Twitter to spread the vile “Pizzagate” story that ended up with an armed confrontation at a Washington pizza restaurant.

Twitter does have a set of rules for its users. And while it bans “threats, harassment and hateful conduct,” noticeably absent is any prohibition against lies and propaganda.

Twitter–and other companies in the information industry–simply cannot have one code of ethics for internal operations that stresses the importance of honesty and respect while it provides an instrument that others use to attack the very values that the company stands for. Similarly, a company that operates within the United States cannot disregard the fact that it provides people with a tool to undermine the core values of the society. Given that, it seems to me that Twitter should ban anyone whose tweets can be shown to be demonstrably false. At the top of the list, of course, is Donald Trump.

Companies in the information industry must begin to accept some responsibility for the harm that their products lead to–especially when that harm can be realistically predicted.