Ask a chatbot which fashion brands are sustainable and the answer depends less on environmental records than on reputation and phrasing. A BoF analysis of ChatGPT, Claude, and Gemini found that a brand’s online image outweighs its actual performance in what the models recommend.
The finding matters because AI assistants are becoming the first stop for curious shoppers. Consumers increasingly ask a model before a search engine, treating the generated answer as a neutral summary rather than a weighted inference.
For brands, the practical takeaway is that sustainability communication is now a form of algorithmic optimization. For shoppers, the lesson is to treat the chatbot reply as a lead, not a verdict, and to chase the underlying evidence.
The technology will improve as models gain access to structured data and verification systems. Until then, the most honest answer an AI can give about fashion’s sustainability may be a question of its own.
Prompt construction shifts the outcome too. The same brand can appear responsible or reckless depending on how the question is framed, a volatility that makes the answers feel more like horoscopes than audits.
