I don’t want AI in my online search. What options do I have?

In the recent announcement made by Google, the tech giant unveiled a new artificial intelligence (AI) policy specifically tailored for schools that utilize Google apps. This policy introduces new settings that allow schools to choose whether to opt-in or opt-out of AI products for their student accounts. For those who prefer to completely shun AI in the classroom, Google also offers the option to turn off search entirely.

However, speaking to The New York Times, Google's vice president for education, Chris Phillips, pointed out that "search and AI are deeply integrated", leading to headlines suggesting that there's now no way to use Google search without AI involvement.

The ubiquity of AI in various aspects of life has consequently sparked growing resistance from individuals who have personal, ethical, or environmental objections to AI. Nonetheless, the necessity to search for information on the internet remains, prompting the question: is it truly possible to search the web without AI? To provide a definitive answer, it is essential to delve into the workings of a web search engine and the role of AI in this process.

Web search engines, like Google, are complex tools that involve numerous components working behind the scenes to deliver search results almost instantaneously. These components include crawling, indexing, spam detection, ranking, user interactions, and evaluation. All these functions rely on algorithms – essentially sets of instructions for computers.

The process of harvesting documents from the web, breaking them down into structured data, filtering out irrelevant content, ordering search results based on relevance, assisting users in formulating queries and consuming search output, and measuring search quality improvements are constantly refined by developers and informational retrieval researchers.

AI systems, particularly machine learning and neural networks, play a significant role in various aspects of web search engines. However, it is important to differentiate between different types of AI. While all components of web search engines likely involve some form of machine learning, the focus here is on generative AI – systems trained on vast amounts of data and capable of generating new content based on user prompts.

This is the technology behind Google's AI Mode and AI Overviews, which provide concise summaries and a novel way of interacting with web content.

Despite the convenience offered by generative AI in search engines, there are notable drawbacks. Generative AI models are built on extensive training data, but they are not updated as frequently as the web search index itself. Consequently, links within AI-generated summaries may become outdated or inaccurate. Additionally, searchers who rely heavily on AI-generated overviews may inadvertently access fewer sources, leading to potential misinformation.

Overreliance on AI summaries can also result in users trusting answers without verifying their accuracy.

At present, commercial web search engines such as Google increasingly incorporate AI-generated summaries into search results. Regrettably, there is no official switch available to disable AI Overviews and other generative AI features in Google search. However, an indirect workaround exists: appending "&udm=14" to the search request link directs search results to the "Web" tab, which excludes AI-generated summaries and snippets.

This method requires manually adding the parameter in the address bar every time a search is performed. To streamline this process, browser plugins and automation tools have been developed for browsers like Mozilla Firefox and Chromium-based browsers such as Brave, Chrome, and Opera on desktop devices.

While Google search remains the most widely used search engine, numerous alternative search engines are gaining prominence, such as DuckDuckGo, Brave Search, and Ecosia. These platforms offer "non-AI modes" that deactivate AI summaries by default and, in some cases, provide the option to filter out AI-generated images. In summary, while the presence of AI in search engines cannot be entirely avoided, there are steps that users can take to limit its impact, including utilizing alternative search engines that prioritize non-AI results.