This bot was brought to you by

Hey, it’s Mark. In California, it’s now illegal to operate a chatbot for commercial or political purposes without disclosing those intentions. The law, which took effect last week, is a symbolic step toward fighting automated disinformation, but it’s unlikely to have an immediate impact in any but the most banal of cases.

That’s because no one is really expected to enforce the law. The state doesn’t have the resources or technical sophistication to police bot activity, and the companies that operate internet platforms were spared from taking responsibility. As the bill wended its way through the legislature, key provisions fell away, leaving those corporations mostly in the clear. The hollowness of the final version highlights a decades-long dilemma over how to regulate online platforms, an issue that has only grown in importance as control of the internet’s most populated destinations is held by just a few companies.