AI-Generated Content Raises Concerns After Appearing in Google News Results

Reviewed By Ayaz Khan. Edited by Asim BN.

Notes: This updated post was improved with the assistance of AI tools and reviewed, fact-checked, and published by a human editor. The initial version was human-written.


———

404 Media’s investigation identified a series of AI-generated articles that appeared in Google News results despite existing policies designed to limit low-quality or automated material. The report documented cases where websites published AI-produced stories that closely resembled legitimate reporting, including articles that reused content from established publications or featured images with visible watermarks. Some of these pieces still surfaced in Google News, raising questions about the effectiveness of current ranking signals in filtering synthetic or repurposed content.

Google Responds to 404 Media’s Findings

In response to 404 Media’s report, Danny Sullivan, Google’s Search Liaison, publicly addressed the issue in posts on X, Mastodon, and Bluesky, clarifying that Google News is not “boosting” AI-generated content.

According to Sullivan, these filters expressly ask “our systems to ignore the regular relevance ranking” and “simply show the latest content in descending order,” which may surface lower-quality results not normally prioritized. He emphasized that Google’s focus “is on the quality of content, [not its method of] production,” reiterating that mass production of low-quality AI content remains against Google’s policies.

Photo: SumUp / Unsplash

Editing Notes:

Multiple attempts were made to contact 404 Media to understand their position, but no clear response was received. Digital Information World acknowledges that portions of this report drew on 404 Media’s reporting. We apologize sincerely for any unintended use of their content without explicit permission. This acknowledgment is strictly for source reference and does not imply approval or endorsement of any other content or positions published by 404 Media. Our intent is to maintain accurate and fair reporting while upholding ethical and responsible journalistic practices.

We apologize for any inconvenience caused by unclear phrasing in earlier versions of this post.

August 4, 2025 Update: The original 2024 article was revised to improve readability.

October–November 2025 Update: Added additional details from Google Search Liaison Danny Sullivan.

December 2025 Update: Restructured post with redundant parts removed.

(Update January/
February 2026) Fact Check: Publishing Frequency and 404 Media AI Assessment

404 Media, in a January 2024 article titled “We Need Your Email Address,” stated that Hura Anwar publishes articles “roughly every six minutes” and “all day every day.” To assess this claim fairly, Hura Anwar’s public publishing activity between 12–26 January 2024 was reviewed. During this period, average posting intervals were consistently measured in hours rather than minutes, and the data shows multiple non-publishing days, contradicting the characterization of continuous “all day every day” output. At no point did the observed activity approach a six-minute publishing interval. The claim that Hura Anwar publishes every six minutes and all day, every day is false and not supported by observed data.

Additionally, the original version of the referenced article was evaluated using multiple AI-detection tools, including Originality AI, GPTZero, and Copyleaks, all of which indicated high confidence (around 99%) that the text was human-written, suggesting it was not generated by AI. Observations from these analyses show that 404 Media’s assessment of AI-generated content, AI plagiarism, or AI spam does not fully align with recognized methodologies, emphasizing the importance of careful scrutiny before drawing conclusions from their reporting.

404 Media’s claim that the article is “completely illegible” is incorrect. DIW acknowledges minor readability issues in the initial version, but independent assessments indicate that the article’s meaning and facts are accessible despite stylistic imperfections. The initial version was evaluated using multiple readability tools: Hemingway App rated it at Readability Grade 9, which is considered good; WebFX Readable measured an average reading ease of 58.8 out of 100, indicating the content is understandable by 15–16-year-olds and comparable to 404 Media’s own original articles; and Originality AI’s Readability Checker gave the article a Flesch-Kincaid Ease Score of 59.4, within the ideal range of 45–60. These results indicate that the article meets standard readability levels and contradict 404 Media’s characterization.

404 Media’s claims regarding this post are not supported by public records and independent assessments.

(Update March 2026): Image refreshed.

Read next: Gaza’s 7-Day Internet Blackout Is The Longest Outage Yet, Web Monitors Confirm
Previous Post Next Post