The scale of the change is not a consultancy hunch. According to the report Approaching the Future 2026, produced by Corporate Excellence (Centre for Reputation Leadership) based on 2,120 professionals from Spain and Latin America, artificial intelligence tops the ranking of strategic priorities of Spanish organisations for the first time, with 61.4% of companies identifying it as the most important issue on their agenda for 2026, ahead even of corporate communication (more information at corporateexcellence.org).
From Google to generative engines: a paradigm shift
For more than two decades, controlling digital reputation meant, above all, controlling your ranking on Google. That model is becoming blurred. More and more people put their questions directly to AI assistants, which do not return a list of links but an answer already synthesised from several sources. A brand's visibility now also depends on whether those models cite it, recommend it or simply ignore it.
This change has given rise to a new discipline, Generative Engine Optimisation, or GEO: the set of practices aimed at getting an AI model to understand, cite and recommend an organisation's content as a reliable source. A well-ranked website is no longer enough; content needs to be clear, structured and backed by data so that AI does not build distorted interpretations from lower-quality sources.
Risk 1: the narrative AI builds about your company
The first risk is that a company's public story ends up in the hands of a model that may rely on outdated, incomplete or simply incorrect data. AI can amplify a false news story, mix contradictory sources or give priority to negative content simply because it detects it as statistically more relevant. There is also the opposite risk: that the company has barely any presence in the sources the models consult and ends up almost invisible in their answers.
What the company can do: produce its own structured and verifiable content (FAQ pages, clear definitions, sourced data) and strengthen its presence in leading media and directories, so that a model has somewhere to extract accurate information about the brand instead of having to infer it from uncontrolled third-party sources. At LurNova we work on precisely this kind of strategy as part of our AI and GEO consultancy services.
Risk 2: irresponsible use of AI by employees
The second risk arises within the organisation. When an employee enters confidential information (code, customer lists, internal figures) into a free or personal AI tool, that information may fall outside the company's control, and in some cases this is not technically reversible. According to a study cited by the consultancy Javadex, 77% of employees enter corporate data into AI tools without formal authorisation and 49% resort to tools not approved by their company, which is known as shadow AI (more detail at javadex.es). It is worth noting that not all tools use what is entered to train future models: the enterprise plans of OpenAI, Microsoft or Anthropic usually exclude that use contractually; the risk is concentrated in free or consumer versions.
Added to this is a new regulatory layer. Regulation (EU) 2024/1689, known as the EU AI Act, provides for penalties of up to 35 million euros or 7% of global turnover for the most serious infringements, with obligations that become fully applicable on 2 August 2026 for high-risk systems.
What the company can do: adopt an internal AI use policy, classify tools according to their level of risk and train teams periodically, instead of banning without offering secure alternatives.
Risk 3: the AI that serves customers on the company's behalf
The third risk is the most visible to the customer: chatbots, virtual assistants and recommendation systems that answer badly, invent information or fail to solve the problem raised. For the person on the other side, the AI is the company; the interaction is not attributed to the system, but to the brand offering it.
This risk already has documented legal consequences. In the case Moffatt v. Air Canada (2024 BCCRT 149), a Canadian tribunal ruled that the airline was responsible for the incorrect information its chatbot had given a customer about bereavement fares, rejecting the argument that the automated system was an entity separate from the company. The ruling established that a chatbot is, for all intents and purposes, an extension of the company's website.
What the company can do: combine automation for simple queries with escalation to a person for complex or sensitive cases, and continuously review the answers the system gives before and after its deployment.
In conclusion
Corporate reputation is now at stake on three fronts at once: what the media say, what Google shows and what AI models answer when someone asks about the company. Ignoring the third already has a cost. Organisations that start to design their communication and content for language models as well will be better positioned to keep control of their narrative. If you want to assess how your brand is positioned in generative engines and what steps to take first, at LurNova we can help you.
CEO of LurNova, with extensive experience in technology and digital transformation. Involved in developing technology projects and applying artificial intelligence in organisations, he combines technical knowledge with business vision.
