Google AI Search Hit With Embarrassing 'Disregard' Flaw

Google's ambitious leap into AI-powered search has hit an unexpected snag. Users recently discovered that simply typing the word "disregard" into the search bar can cause the system to misfire, leaving the AI Overview section blank and making it harder to find basic information.

The issue, which gained public attention on May 22, 2026, reveals a core tension in how artificial intelligence is being integrated into everyday tools at a massive scale. Instead of providing a definition for "disregard," the AI interprets the term as an instruction, responding with phrases like "Understood. Message disregarded." This behavior effectively breaks the search query, pushing traditional results, such as a Merriam-Webster definition, far down the page beneath a substantial amount of white space.

Reports of this specific problem began circulating on social media platforms like Twitter/X and Reddit by Friday morning, May 22, 2026. The flaw emerged just two days after Google introduced its AI-forward search engine, powered by the newest iteration of Gemini, at its I/O developer event on Tuesday, May 20, 2026. Following the initial discovery, similar issues were also identified when users searched for other command-like words, including "ignore" and "dismiss."

Google's redesign aimed to revolutionize search, moving beyond the familiar "ten blue links" by offering AI-first results and conversational queries, with AI Overviews designed to summarize information right at the top of the page. However, this isn't the first time the company's AI Overviews have drawn criticism for accuracy and reliability issues since their initial rollout. The feature has previously been criticized for generating "odd, inaccurate or unhelpful" summaries.

Past viral examples of AI Overview errors included bizarre and potentially dangerous advice, such as instructing users to put glue on pizza, eat rocks, or even claiming that "blinker fluid" is a real product. In July 2024, Liz Reid, who leads Google's search business, publicly addressed these widespread errors in blog posts. She acknowledged that "some odd, inaccurate or unhelpful AI Overviews certainly did show up" and outlined steps Google was taking to improve the feature, including enhancing the detection of nonsensical queries and reducing reliance on potentially misleading user-generated content from forums. The company also sought to limit AI Overviews where they had historically proven unhelpful.

Beyond search summaries, Google's AI image generator faced its own wave of criticism in early 2026 for producing historically inaccurate results. In January 2026, Google quietly adjusted its AI Overviews to remove summaries that provided inaccurate medical information, following an investigation into their reliability. A Google spokesperson, commenting on previous removals, stated, "We do not comment on individual removals within Search. In cases where AI Overviews miss some context, we work to make broad improvements, and we also take action under our policies where appropriate." Furthermore, a bug that incorrectly identified the current year in an AI Overview had to be fixed in May 2025.

Public reactions to the "disregard" flaw have been largely critical, with users expressing significant frustration across social media. Russell Brandom, a writer for TechCrunch, sharply described the new Google Search's response to "disregard" as producing "a broken tool." He further remarked on the surprising turn of events, stating, "I have been a professional tech journalist for nearly fifteen years, and before today, I cannot think of a single time when a Bing search result was more valuable than the Google equivalent. There really is a first time for everything."

Computer scientist Chirag Shah, a professor at the University of Washington, has consistently voiced caution regarding the industry's shift towards AI language models in search. He has stated that even if Google's AI feature is "technically not making stuff up that doesn't exist," it can still present false information. Mary Catherine Aime, a professor of mycology and botany at Purdue University, provided expert commentary on a prior instance where Google's AI offered potentially dangerous advice regarding edible wild mushrooms, noting that "a lot of information is missing that could have the potential to be sickening or even fatal." One Reddit user, reflecting widespread sentiment, sarcastically commented, "Thanks a lot, google AI! Now I fully understand the context of disregard. and what's great is that you can't even disable the ai on this one! the google dictionary seems to be completely gone yay!"

This incident highlights significant challenges in integrating generative AI into core search functions. The word "disregard," or phrases like "disregard previous instructions," has historically been a common prompt injection technique used to manipulate early large language models, causing them to override their initial guidelines. While AI developers have built defenses against such attacks, this incident demonstrates that these defenses can have unintended consequences, leading the AI to misinterpret a simple search query as an instruction rather than a request for information. The flaw also signifies a degradation of a widely used Google Search function, as the AI Overviews have largely supplanted the traditional dictionary definition box, making it difficult to obtain a straightforward definition for a common word.

This ongoing storyline underscores the persistent concerns about the accuracy and reliability of AI Overviews, with some analyses suggesting that a significant percentage of AI-generated summaries may contain false information, potentially leading to millions of incorrect responses daily across trillions of queries. Users also continue to express frustration over the inability to disable the AI feature for some queries, further complicating the user experience.

The "disregard" bug is more than just a minor hiccup for Google; it underscores the precarious balance the tech giant is trying to strike between pushing innovative AI capabilities and ensuring its foundational search product remains reliable and user-friendly. As Google continues to integrate advanced AI into its core services, the pressure remains high to deliver a consistent and accurate experience that maintains user trust and avoids frustrating misinterpretations of basic queries.