In the context of internationalisation, intercultural communication is essential for building trust-based relationships and effectively promoting a brand and its products and services. Artificial intelligence offers powerful tools to support these activities, but it can also pose significant risks if it is not used consciously and responsibly. One particularly relevant example concerns image generation: a study conducted by Bloomberg, for instance, analysed images created using the Stable Diffusion model and highlighted how its outputs tended to reinforce racial and gender stereotypes. A manager, for example, might typically be portrayed as white, male and with Anglo-Saxon features, while a warehouse worker might be depicted as shorter and with Latin American features.
Claude: supporting more stereotype-aware prompting
Let us consider an example of how stereotypes can be avoided in international communication. If we ask ChatGPT to generate an image depicting a young Chinese woman waiting to receive a product purchased online, the result may include exaggerated features, traditional clothing presented without appropriate context, and a background designed to evoke cultural clichés. Such a representation is not only unrealistic but also risks perpetuating distorted perceptions of a rich and complex culture.
To address this type of distortion, Claude, with its advanced language models, can be used to reformulate the prompt in a more neutral and respectful way. Subsequently, Ideogram, an image-generation tool, can be used to create a new image depicting a young Chinese woman in a modern urban setting, wearing contemporary clothing and portrayed in a more natural, non-stereotypical way. This approach demonstrates how careful prompt formulation and the use of appropriate tools can lead to more accurate and respectful representations.
ChatGPT for countering stereotypes in international communication
ChatGPT can also provide effective support in adapting communication content to the different levels of cultural familiarity and value systems that characterise international markets. In an internationalisation process, one of the most delicate aspects concerns the semantic and symbolic adaptation of concepts that are deeply rooted in the Italian business vocabulary – such as “Made in Italy”, “craftsmanship”, “short supply chain”, “design” and “manufacturing excellence” – which, despite having a well-established meaning and value in Italy, do not necessarily retain the same effectiveness or significance in other cultural contexts.
With ChatGPT, it is possible to carry out preliminary checks that can be valuable in refining a message. For example, the model can be asked whether, in a particular country, terms such as “Carrara marble” or “barrique” are widely recognised and whether their evocative value is understood. In geographically distant or non-specialist markets, their impact cannot be taken for granted: what in Italy evokes tradition, prestige and a strong connection to local heritage may elsewhere be perceived simply as a building material or a generic winemaking technique.
Similarly, it is possible to assess whether words such as “luxury” or “tradition” carry positive connotations or whether, in certain contexts, they may be perceived as outdated or elitist. In some Asian markets, for example, the concept of “tradition” needs to be carefully contextualised, as it may take on different meanings depending on the degree of modernisation perceived or desired by the local audience.
Another strategic use of ChatGPT lies in the possibility of requesting not simply a literal translation, but a genuine cultural adaptation of the message. Instead of asking, “How do you translate artigianalità into English?”, one might ask, “How can the deeper value of Italian craftsmanship be communicated to a US audience in a way that conveys authenticity, uniqueness and expertise without sounding old-fashioned?” The result is often not a single word, but an entire semantic field that can then help guide the communication and storytelling strategy.
In addition, ChatGPT can be used to explore how a foreign audience might respond to a campaign built around specific Italian values. For example, if a visual communication campaign is being developed around Renaissance aesthetics, the model can be asked how a Japanese or North American target audience might perceive it. The analysis may suggest cultural interest and enthusiasm, but also a sense of distance or misunderstanding if the concept is not accompanied by appropriate contextualisation.
Finally, even short and apparently neutral messages – such as a slogan or tagline in English – can be tested in advance to determine whether they retain semantic coherence or risk sounding generic, ambiguous or insufficiently distinctive. A phrase such as “From Italy with Craft” may appear effective from an Italian perspective, but could be perceived as unclear or overly generic in the UK or US market.
Across all these applications, ChatGPT does not replace a company’s strategic and cultural expertise, but rather acts as an accelerator of awareness and a tool for helping to prevent communication mistakes. The added value of AI lies precisely in the possibility of rapidly carrying out preliminary assessments of the cultural compatibility of content intended for international markets, thereby supporting communication that is genuinely inclusive, respectful and targeted. It is through this integration of human judgement and AI-assisted analysis that artificial intelligence can become an effective tool in support of internationalisation.
Is it mandatory to disclose the use of Artificial Intelligence?
An increasing number of misleading or manipulated images and videos are being created using Artificial Intelligence. The European Union’s AI Act, which entered into force on 1 August 2024, establishes specific transparency obligations for certain AI systems and AI-generated or manipulated content. The transparency requirements set out in Article 50 became applicable on 2 August 2026.
This does not mean that every use of Artificial Intelligence must automatically be disclosed. However, providers of certain generative AI systems are required to enable AI-generated or manipulated content to be identifiable in machine-readable form. In addition, deployers of AI systems that generate or manipulate deepfake image, audio or video content must disclose that the content has been artificially generated or manipulated, subject to specific exceptions provided for by the legislation. Certain AI-generated texts concerning matters of public interest may also be subject to disclosure requirements when published without human review or editorial control.
Artificial intelligence therefore offers significant opportunities for international communication, extending far beyond support for text translation, but it is essential to use it consciously and responsibly. Avoiding ethnic and cultural stereotypes, ensuring appropriate transparency in the use of AI, and adapting content to different cultural contexts are fundamental aspects of communication that is respectful and attentive and, for this very reason, more effective.