The contexts in which this new “ecosystem” finds application are numerous: a business phone call or a video call in a foreign language, but also a trip abroad or any situation where you need to communicate with someone speaking a different language. Today, real-time translations powered by artificial intelligence are accessible through a wide range of tools, from everyday devices such as earbuds and smartphones to the chatbots we have all become very familiar with. A true ecosystem, indeed, that is breaking down language barriers, although there is still a way to go before reaching perfection. We are certainly not yet in the scenario imagined by Douglas Adams with his “Babel Fish,” which allows understanding any language in the galaxy. Still, it is also true—and this is no small detail—that all major tech players are active in this field, from Google to Microsoft, from Meta to Apple, and including the main Asian manufacturers.
Take Apple, for example: with the latest version of iOS, the Live Translation feature has arrived on AirPods, allowing real-time translation of face-to-face conversations, with the translated voice delivered directly through the earbuds. Just a gesture or a Siri command, and the invisible technology Apple loves so much comes into play once again, leveraging AI capabilities. Meta, for its part, has brought voice translation directly into the second-generation Ray-Ban smart glasses, confirming the vision of an augmented reality that frees us from traditional screens: the integrated smart assistant interprets what the user hears or says, delivering the processed output through the device’s mini-speakers or transcribing the text in the Meta AI app.
Another “hot front” for AI-driven translations is messaging platforms, with WhatsApp being the most popular example. Here, the transformation of entire conversations or individual messages occurs directly on the user’s device, preserving end-to-end encryption.
AI translation has thus become a structural feature of smartphones. Samsung paved the way by equipping its Galaxy devices (from the S24 Series onward) with simultaneous interpreters for face-to-face conversations and phone calls. Google responded with its Voice Translate integrated into the Pixel 10, translating calls while preserving the original voice tone thanks to on-device processing. Chinese companies such as Xiaomi, Oppo, and especially Honor have followed suit, with the Magic V5 integrating a full large speech model directly on the device to enable automatic translation from the phone itself, independent of the cloud.
The competition among Big Tech extends to collaboration platforms used by businesses and professionals, with Google Meet and Microsoft Teams leveraging proprietary AI agents or third-party software to provide simultaneous translation with synthetic voices, making these tools increasingly useful and functional in meetings and business contexts. In this perspective, the latest innovation from the German company DeepL is noteworthy: DeepL Voice, a technology (already integrated into Teams and Zoom Meetings) that allows users to speak in their own language during a call while showing real-time translated subtitles to other participants and translating live dialogues via smartphone.
The most recent announcements from Google and OpenAI, finally, have shifted attention back to chatbots and generative AI models. With ChatGPT Translate, OpenAI adds automatic translation to its galaxy of services with a distinctive feature: the literally translated text serves as a starting point to rework the output according to context, creating “tailored” content. TranslateGemma, on the other hand, is Google’s latest offensive: not a consumer service “tout court” but a family of “open-weight” models (pre-trained weights are released for local download and custom execution) designed for developers and researchers. Built on Gemma 3, available in three sizes (4, 12, and 27 billion parameters) and capable of translating up to 55 languages, these models are trained in two phases to produce more natural and contextual translations. What makes this announcement significant, perhaps, is Google’s intent to bring the translation tools challenge to the infrastructure level.
