SIRI-OUSLY 2.0: What Artificial Intelligence Reveals about the First Amendment

Innovation and Economic Growth, Cloud Computing, Networks, the Internet, and Cloud Computing and Artificial Intelligence

Article Snapshot


Margot Kaminski, Toni M. Massaro and Helen Norton


Minnesota Law Review, Vol. 101, No. 6, pp. 2481-2525, 2017


Machines that can actually think are referred to as strong Artificial intelligence (AI). The First Amendment might protect speech by strong AI. Courts focused on the value of speech to listeners and the need to constrain government power will be sympathetic to this view.

Policy Relevance

AI free speech could spur courts to clarify the value and limits of free expression.

Main Points

  • “Positive” theories of free speech that focus on its value to democracy note the value of free speech to listeners, and speech by AI could contribute to this value.
  • The Supreme Court’s First Amendment cases support a “negative” view of free speech, the idea that government regulation of speech is not permitted without strong justification.
    • Regulation of a few traditional categories of troublesome speech is permitted.
    • The Court questions the government’s competence to decide whether speech is beneficial or harmful.
    • Government officials might use power to control speech to protect its own interests.
  • The First Amendment states that the government “shall make no law” to suppress speech.
  • Protection of AI free speech could protect AI-to-AI speech as well as AI-to-human speech.
  • Some might argue that AI should not have free speech rights, because they are not persons with emotions or moral autonomy.
  • Animals would probably not qualify for free speech rights even if free speech rights are granted to AI, because AI is evolving more rapidly, and is human-centered and designed; most of what animals communicate would not be “speech” under the first amendment.
  • Courts would need to limit the rights of AI to make defamatory statements, even when AI could not be said to have acted carelessly or intentionally in making such statements.
  • The owners of AI could be sued for the consequences of harmful AI speech.


Get The Article

Find the full article online

Search for Full Article