This article was compiled by the editorial desk, drawing on public statements, company websites, and direct testing of the tool mentioned in the original report.

In a recent thread on X, Jake Ward, founder of the UK-based SEO agency Content Growth, openly described what he called an "SEO heist" that diverted 3.6 million visits from a competitor. Ward detailed how he obtained the competitor's sitemap, converted its URLs into article titles, and then used an AI tool to generate 1,800 articles based on those titles. The posts quickly drew backlash from users who accused him of plagiarism and questioned the ethics of his approach.

Ward's revelation comes at a time when AI-generated content is becoming increasingly prevalent online, raising concerns about the quality and originality of information available to internet users. The practice he described involves repurposing existing content through AI algorithms, a method that some see as a way to game search engine rankings rather than provide genuine value to readers.

The AI tool Ward used is called Byword, a company he co-founded earlier this year. Byword's website openly promotes a feature designed to help users avoid detection by AI content detectors. The site claims that enabling this feature instructs Byword to vary word and sentence structure in a way that differs from other AI content generators, making its output more difficult to identify as machine-generated.

AI Detection Challenges

Byword is built on OpenAI's GPT-4 model and offers various pricing tiers, from $5 per article to $2,499 per month for unlimited articles and a dedicated server. The company's goal, as stated on its website, is to reduce the effort required to generate content and allow users to spend more time on strategy. However, the effectiveness of such tools in producing content that passes the "smell test" remains uncertain.

Testing Byword for this article revealed notable shortcomings. A request for a 1,600-word blog post on the topic "first exoplanet discovered" produced text that only mentioned the first exoplanet about three-quarters of the way through, after a lengthy preamble about astronomical research. The article also featured a structure of SEO-friendly headers that seemed designed to appeal to search algorithms rather than guide readers.

Another test, based on the keywords "melting eggs," generated an 1,800-word article titled "The Art of Melting Eggs: A Culinary Delight," which provided nonsensical instructions for melting eggs over low heat for a "luxurious breakfast." A separate 1,500-word piece on "investing in Tesla" made no mention of CEO Elon Musk, a notable omission given his influence on the company's valuation.

Industry Context

Ward's actions are part of a broader trend of publishers using AI to generate content, often with mixed results. Earlier this year, BuzzFeed faced criticism for publishing AI-generated articles that contained awkward language and copied phrases. Some content farms have been caught leaving the phrase "as an AI language model" in their output, inadvertently revealing the use of AI. In a more extreme case, Sports Illustrated was found to have published articles by authors who were entirely AI-generated.

The practice of using AI to rewrite competitors' content raises legal and ethical questions. Proving that AI was used to repurpose content is difficult, as AI detectors are not yet reliable enough to distinguish between human and machine-generated text. Even OpenAI, the maker of ChatGPT, has admitted that its tools cannot reliably detect AI-generated content, complicating efforts to address plagiarism.

Ward's "heist" highlights the growing pressure on companies to rank high in Google search results, often at the expense of content quality. As AI tools become more accessible, the line between original journalism and algorithmically generated filler continues to blur, leaving readers to navigate an increasingly noisy digital landscape.