Sustainability does not become credible because we communicate it better.

Sustainability does not become credible because we communicate it better. It becomes credible when what a company says matches what people see, experience and recognise.

A systematic literature review on the authenticity of corporate social responsibility analysed 46 peer-reviewed studies. Among the factors contributing to perceived authenticity are consistency between values and actions, credible motivations and continuity in stakeholder engagement.

For those working in communication, the question changes. It is not enough to ask “How do we communicate our sustainability?”. We should ask: “Is what we communicate consistent with what our stakeholders experience of the company?”.

Authenticity is not just an attribute of the message: the study describes it as something built through alignment, consistency and interaction with stakeholders. Employees and consumers become part of the process through which sustainability is recognised, or questioned.

And today there is a new interlocutor to consider: AI. When we ask ChatGPT, Gemini, Claude or Perplexity to tell us about a company, the answer does not come only from its sustainability report or its website. LLMs rebuild a representation of the brand from the information they manage to retrieve and connect.

This is where sustainability reputation and AI Reputation begin to meet. A company can state very clear values and commitments on its own channels, while the representation returned by AI may be incomplete, different or not backed by sufficiently recognisable evidence.

The question becomes twofold: do our stakeholders recognise what we say as authentic? And is AI building the same representation of the company?

For corporate communication, this means moving from simply producing narratives to managing consistency between reality, evidence, stakeholder perception and AI perception. Today what matters is not only what a company says it is, but what it can demonstrate, make recognisable and make understandable, to AI as well.