CEOs Around the World Are Getting the Same Generic Advice From AI — Researchers Call It 'Trendslop'

Harvard Business Review reports that CEOs around the world are receiving strikingly similar business guidance from large language models, with researchers describing the phenomenon as 'trendslop.' The findings were discussed in a recent review of AI-generated strategic advice across multiple models, including GPT, Claude, Gemini and Grok.

The research suggests that when asked tens of thousands of times for business strategy input, AI systems repeatedly produced near-identical recommendations regardless of industry, context or level of detail.

The advice tended to cluster around themes such as 'differentiate,' 'collaborate,' 'think long term' and 'augment existing capabilities,' raising questions about how much real-world specificity these tools are offering senior decision-makers.

Recently, corporate reliance on AI tools for strategy support, brainstorming and executive decision-making has grown. While these systems are often marketed as adaptive and context-aware, the findings suggest a strong tendency towards safe, widely repeated business language.

CEOs and the 'Trendslop' Problem

The research cited in the Harvard Business Review analysis involved repeated prompting of major AI models tens of thousands of times, with researchers testing whether outputs would change based on industry type, wording shifts or incentive structures.

According to the findings, even when prompts were altered or models were effectively 'rewarded' for variation, the core advice remained largely unchanged. The pattern was consistent across different systems, including GPT-style models, Claude, Gemini and Grok.

Researchers noted that outputs tended to flatten into familiar corporate language often seen in presentations, consultancy reports or leadership think pieces. In other words, the kind of phrasing that sounds strategic but often avoids specifics.

This is where the term 'trendslop' emerged. It refers to AI-generated responses that sound widely accepted and popular but are often vague and not tailored to specific situations. These outputs tend to rely on familiar buzzwords such as innovation, transformation and collaboration without offering concrete, practical steps.

Researchers say this occurs because large language models are trained on vast amounts of internet text, including blogs, social media posts, talks and corporate documents. As this training data contains repeated ideas and common phrasing, the systems learn to reproduce those patterns. As a result, the models often reflect what sounds typical or popular online rather than generating advice grounded in specific real-world context.

Average Advice for Decision-Making Equals Business Problems

The study's main concern is not that AI produces incorrect answers, but that it often delivers average, middle-of-the-road advice. Instead of engaging with the specifics of a situation or challenging assumptions, it tends to fall back on widely accepted ideas.

Researchers warn this could pose a problem for business leaders. If companies across different industries are receiving similar guidance from AI, it may become harder for them to develop distinctive strategies and stand out from competitors.

One way to understand this is that AI does not 'think' like a human. It predicts the most likely next words based on patterns in its training data. As a result, it often repeats ideas that are common online rather than generating advice tailored to a specific real-world context.

This has prompted mixed reactions. Some argue AI is useful because it reinforces established business principles. Others warn it could encourage uniform thinking, which may be risky in competitive industries where originality matters.

Overall, the study is not suggesting AI cannot assist with strategy. Rather, it highlights the need for caution in treating polished, confident-sounding responses as inherently original or fully tailored.