•  
  •  
 

Abstract

Congress passed Section 230 of the Communications Decency Act of 1996 in response to two civil cases from New York that split on the issue of immunity for online platforms from civil liability for content posted on their websites by third parties. Congress was concerned that, without immunity, online platforms would not survive, thereby hampering the internet’s development. Congress was right; commentators have hailed Section 230 as the backbone of the internet. But as time has passed, dissidents have argued that Section 230 provides broad immunity outside the bounds of what its drafters anticipated. With the emergence of generative artificial intelligence, Section 230 is back in the news. Congress, President Joe Biden, law firms, and law professors have begun discussing in executive orders, hearings, and law review articles whether Section 230 immunity applies to generative artificial intelligence. All are concerned about the impact that artificial intelligence will have on our society. The issue is still novel, so circuit courts have not had the opportunity to decide whether Section 230 immunity applies to generative artificial intelligence.

This Article argues that Section 230 immunizes generative artificial intelligence but also that it should not. That is because broad immunity for generative artificial intelligence will leave those harmed by it without recourse. Instead of granting broad immunity, Congress should base immunity on the generative artificial intelligence company’s effort to prevent the kinds of harm that AI produces. This Article proposes that Congress should borrow from other areas of the law, where Congress has conditioned immunity on proactively preventing harms that the potential tortfeasor is aware of, to fashion immunity for generative artificial intelligence only under specific circumstances.

Download

Share

COinS