Are your online browsing habits changing? If so, the answer might lie with the answer engine.
It is not often that a dictionary takes someone to court. But in September 2025, the Merriam-Webster Dictionary and its parent company, Encyclopedia Britannica Inc, did exactly that. They accused an answer engine called Perplexity AI of unlawfully copying their content and diminishing their monetisation and financial revenue by cannibalising web traffic away from their websites. The lawsuit also alleged that the answer engine had violated Merriam-Webster’s trademark by attributing AI hallucinations to the dictionary. It is the first lawsuit initiated by a dictionary in nearly half a century, when Oxford English Dictionary had sought (and obtained) an injunction against Pergamon Press for using its name in the title of the Pergamon Oxford Perfect Spelling Dictionary.
An answer engine is an online tool that provides AI-generated responses to queries, instead of a list of web results for users to browse themselves. In contrast with a traditional search engine, which acts as an intermediary between users seeking information and web publishers who provide it, answer engines do the digging for the user by copying data from websites and summarising it in the form of a direct answer. Even the world’s largest search engine, Google, is shifting towards the answer engine model by providing “AI Overviews” and “AI Mode” panels above its traditional results.
This shift is disrupting how business is done on the internet. Fewer people are clicking through to original websites and, as seen in the Merriam-Webster lawsuit, zero clicks equate to a drop in ad revenues and subscriptions. Publishing platforms and content creators are finding that their old strategy for getting noticed online, including search engine optimisation (SEO), is losing its power. They are now scrambling to devise answer engine optimisation (AEO) or even large language model optimisation (LLMO). Some are also forming paid partnerships directly with answer engines.
Exploring the history of answer engines takes us back to 1961 when MIT computer scientists, among them Alice Wolf and Carol Chomsky, created a computer program called Baseball, which read natural language questions about baseball games from punch cards and gave direct answers. When the world wide web launched in the mid-1990s, sites such as Ask Jeeves presented themselves as answer engines, followed in 2009 by WolframAlpha.
As today’s answer engines steal traffic from websites and the whole economy of the internet finds itself in flux, language relating to the internet will change too. Users will switch from searching with keywords to asking complex questions in natural, conversational language. Terms such as traffic, hits and clicks are already giving way to content, cited sources and AI mentions. When machines start answering for us, what happens to the joy and serendipity of discovering for ourselves? Will we keep browsing, exploring and surfing the internet, if the spontaneous agency of users is replaced by the singular output of an answer engine?
