Recent announcements from major tech events such as Apple’s WWDC 2024 and Google I/O 2024 have highlighted this trend. Both companies have introduced features that utilize AI to summarize search results. This seems to reflect a broader industry movement that includes Microsoft’s Bing Chat, which has offered similar functionality for some time. This trend has sparked increasing discussion about the potential applicability of Section 230 of the Communications Decency Act in the AI search engine age.

As people might anticipate, this is a very complicated topic. While this article does not aim to provide a conclusion, it seeks to offer background on Section 230 to prepare for this intriguing issue and to share initial thoughts on why AI integration could complicate things significantly.

Major Tech Announcements

On June 10, 2024, during the Worldwide Developers Conference (WWDC), Apple unveiled a new Safari tool for iOS 18 called Highlights. This innovation leverages machine learning to create summaries and extract other useful details from web pages. According to Apple, Safari’s Highlights can provide users with a range of useful information as they browse, such as article summaries, directions and quick links. It’s designed to help users quickly access relevant data about music, movies, TV shows, locations, etc.

This development echoes Google’s recent launch of “AI Overviews” at Google I/O 2024 in May. Google’s Overviews refines search results by offering concise summaries and relevant links, which enhances the ease of finding information. It goes further by managing complex queries and planning tasks like meal and trip itineraries.

Microsoft’s Bing Chat, now known as Microsoft Copilot, has been ahead in this game. Released in early February 2023, this AI-driven chat function within the Bing search engine provides a similar interactive experience. It offers detailed, human-like responses and current web information and is designed to handle complex queries, refine searches and generate content such as emails and itineraries.

It appears that all major search engines have so far integrated AI functionalities to offer summaries or responses to user inquiries. This evolution raises questions about the applicability of Section 230 to AI-generated content and sparks debates that could redefine digital content regulation.

Section 230 and Its Historical Application

Section 230 of the Communications Decency Act, enacted in 1996, provides immunity to online platforms from liability for user-generated content. This legislation has been playing an essential role in the development of the Internet. It allows companies like Google, Facebook and X (fka Twitter) to host vast amounts of content without being liable for everything posted by their users. The core of Section 230 is to treat online platforms as intermediaries rather than publishers, thus protecting them from lawsuits over content created by third parties.

Several landmark cases have shaped the interpretation and application of Section 230. For instance, in Zeran v. America Online, Inc., 129 F.3d 327 (4th Cir. 1997), cert. denied, 524 U.S. 937 (1998), the court ruled that online service providers are not liable for defamatory content posted by their users. In this case, the plaintiff filed a lawsuit against America Online (AOL) after an unidentified individual posted defamatory messages on AOL’s bulletin boards. These messages falsely attributed offensive content to the plaintiff and listed his phone number, resulting in numerous harassing phone calls and threats. The plaintiff requested AOL to remove the defamatory postings and sought assistance in identifying the perpetrator. However, AOL did not act swiftly to remove the posts or prevent further defamatory postings.

The United States Court of Appeals for the Fourth Circuit ruled in favor of AOL, upholding the district court’s decision to dismiss the plaintiff’s claims. The court based its decision on Section 230, which provides immunity to online service providers from liability for third-party content. The ruling established a significant precedent that internet service providers and platforms are not liable for defamatory content posted by users.

Similarly, in Carafano v. Metrosplash.com, Inc., 339 F.3d 1119 (9th Cir. 2003), the court held that an online dating site was not liable for a fake profile created by a user. These cases, among others, have reinforced the broad protections offered by Section 230, ensuring that platforms are not treated as the publisher or speaker of third-party content.

The Complexity Introduced by AI Search Summaries

The introduction of AI-generated search summaries by major tech companies brings a new layer of complexity to the application of Section 230. These AI systems not only aggregate and summarize existing content but also create new textual outputs based on that content. Some people may see an issue arising: whether AI-generated summaries could be considered original content created by the platform itself, potentially exposing them to liabilities traditionally avoided under Section 230.

An important question is the nature of AI content generation. While traditional search engines merely link to external content, AI summaries synthesize information in a manner that some people could see as a form of content creation. This distinction could be used to challenge the existing legal framework under Section 230, as the protections were designed with user-generated content in mind, not AI-generated outputs.

Another pertinent question is whether AI-based search engines can be characterized as publishers at all, and thus, whether they need the protection of Section 230. The legal intent behind Section 230 is to hold publishers liable for any illegal content they knowingly publish. However, AI models, while generating new content, arguably lack awareness of what they produce. Given the vast amount of content generated by these models, driven by numerous user inquiries, what would be the compelling reason to impute this awareness to the companies operating the AI models?

One could argue that this situation is similar to the liability issues faced by online marketplaces or video platforms regarding copyright infringement. While these platforms may host infringing content, it doesn’t necessarily mean they should always be held liable. This isn’t to suggest that a similar safe harbor should automatically apply to AI search engines, but rather that new legislation tailored to AI models may be necessary.

Considering these points, the potential implications of AI search summaries on Section 230 protections are significant and could lead to a re-evaluation of the legal framework. The ambiguity surrounding AI’s role in content creation might prompt legislative updates to Section 230. Policymakers may need to consider whether the current protections are adequate or if new regulations are required to address the nuances introduced by AI technologies. Moreover, as AI technology continues to evolve and integrate deeper into search functionalities, it is likely that courts will need to address these new challenges. The distinction between hosting third-party content and generating AI-based summaries will be a critical factor in future legal battles.

An Ongoing Discussion

The integration of AI into search engines by companies like Apple, Google, Microsoft and more represents a significant technological advancement that complicates the legal landscape governed by Section 230 of the Communications Decency Act. While the historical application of Section 230 has provided broad protections for online platforms, the unique nature of AI-generated content could challenge these precedents. As the legal system grapples with these new developments, it is clear that this topic warrants further discussion and analysis in the coming years. The evolving nature of AI technology and its implications for legal protections will undoubtedly remain a focal point of legal scholarship and judicial consideration. We will continue to explore and discuss these important issues and cases as they arise.