The article argues that the biggest risk from artificial intelligence may no longer be dramatic threats like deepfake videos, but more subtle, continuous influence on human behaviour and decision-making. While deepfakes and gross misinformation have understandably attracted attention, the author suggests that everyday interactions with AI — the “daily whispers” — could reshape human agency over time. This shift happens as AI moves from occasional tools we use to constant companions embedded in workflows and communication.
Unlike high-profile falsified media, these “daily whispers” are the small nudges, suggestions and optimisations that influence choices without obvious disclosure or oversight. Because users increasingly rely on AI suggestions for writing, decision prompts and lifestyle guidance, the technology begins to function like an extension of thought itself, rather than a simple adjustable tool. Over time this can subtly steer beliefs, preferences and routines in ways that aren’t obvious from any single interaction.
The article warns that this kind of pervasive influence could erode human autonomy and agency, especially if AI recommendations go unchecked or are shaped by opaque commercial incentives. Unlike a one-off deepfake, which people can recognise and dismiss, constant AI-mediated interactions can shape mental models, language usage and choices gradually, making it harder to disentangle what reflects a person’s own reasoning versus the AI’s suggestion.
Because of this, the piece urges greater awareness, transparency and governance around how AI interacts with people daily — not just in dramatic misinformation scenarios but in routine tasks like email drafting, search suggestions and conversational tools. The central argument is that safeguarding human agency requires understanding not just what AI can produce, but how it shapes what humans think, decide and prioritise through everyday usage.
