Google’s AI Overviews—a generative AI feature —has triggered an antitrust complaint from independent publishers in the EU, who allege that the feature harms their traffic, revenue, and long-term sustainability.
About Antitrust
- Antitrust refers to laws and regulations that prevent companies from engaging in unfair practices that harm competition, such as monopolies, price-fixing, or abusing market dominance.
- For example, Google faced antitrust investigations for allegedly using its dominance in search and Android to unfairly promote its own services over competitors’, limiting consumer choice.
|
What are Google AI Overviews?
- AI Overviews are AI-generated summaries that appear at the top of Google’s search results page, above the traditional blue links. They synthesise information from multiple sources to quickly answer user queries.
- These overviews may take the form of paragraphs, lists, or tables, and include embedded source links.
About Generative AI
- It refers to artificial intelligence systems that can produce original content—such as text, images, audio, video, or code—based on user prompts or instructions.
- These tools are powered by advanced AI models, such as Large Language Models (LLMs), which form the backbone of many text-based applications like ChatGPT.
- Generative AI is built on deep learning, a branch of machine learning that uses algorithms modeled after the human brain’s neural networks to simulate learning and decision-making.
|
How do AI Overviews work?
- When a query is entered, Google evaluates whether a generative AI response would be helpful.
- It uses a customized version of its AI model, Gemini, combined with a method called Retrieval-Augmented Generation (RAG), which pulls information from the web index in real time.
- The output is a summary backed by top web results, with clickable links for deeper exploration.
Retrieval-Augmented Generation (RAG) is an AI framework that combines the strengths of traditional information retrieval systems (such as search and databases) with the capabilities of generative large language models (LLMs). |
Why are publishers accusing Google?
- Reducing user click-throughs to publisher websites, since users often get their answers directly from the AI summary. This drop leads to lower ad revenue and subscriptions, which are crucial for publishers.
- Google is “misusing web content” by scraping publisher material to train its AI and display summaries—without fair compensation.
- As of May 2024, Google also began placing ads within AI Overviews, thereby monetising content created by others.
- Publishers cannot opt out of AI Overviews without also being removed from Google’s search results entirely—an unfeasible option given Google’s market dominance.