The rising star of Speech-to-text Interpreting : how AI transcription tools impact its sustainability within the sensitive accessibility ecosystem

On the flourishing ground of accessibility regulations and thereby fostered interest of T&I institutions to offer curricula with accessibility focus, speech-to-text interpreting (STTI) has gained considerably ground in academic training and research (Platter 2025, Norberg & Stachl Peier 2018, Eichmeyer-Hell 2020, 2021, 2025). Therefore, unlike in its beginnings where clients had difficulties in finding professionals, particularly considering interlingual and intralingual STTI for the German language, a significant number of specialized STTI professionals has been trained within the last two decades. Nowadays, these professionals are available for potential clients in the public and private sector and can be contacted by registers and expert lists established by STTI associations. As a given reality and like many other professions, speech-to-text (STT) interpreters for the delivery and performance of their services are part of professional networks: the latter being defined as groups of professionals closely working together on the same project, in a shared working context with a high degree of interaction and coordination (Risku et al. 2016). While interaction and collaboration entities, parameters and activities for interpreters collaborating in booth or for interpreters in teams have already been addressed in conference interpreting research (e.g. Nogueira 2022; Hoza 2022, Platter et al. 2025), for STTI the definition of collaboration entities, parameters and activities involved in the organization and implementation of STTI services has not been given much attention so far. Undoubtely, there are collaborative process phases which enable, facilitate and enhance the provision of the service in the broader social and economic context. As interpreting specialisiation, STTI has always been impacted and even defined by the available and feasible technological tools, starting from the earliest approaches of handwriting, (conventional and specialized) keyboard-based and in the later stages speaker-dependent speech-recognition systems. While professionals are aware of the fact, that technology is just one aspect of professional, specialized, and therefore efficient live rendition of spoken into written texts, with the dawn of AI-based, apparently free to use tools which are available 24/7, entities within the broader interpreting network seem at least to be tempted to try or even substitute human STT interpreters in different kind of interpreting settings. Such developments have direct implications for STTI professionals, in terms of assigned interpreting jobs, estimated incomes and rates paid for STTI services, thus impacting mid- and long-term career planning and economic sustainability. In our study we focus on two different settings of STTI provision: one being STTI as public service interpreting setting for a single individual user known to the STT professionals, the other being STTI for an anomymous, heteregoneos user group in the context of live captioning for a private broadcasting provider. In both settings, the STTI service is organised and coordinated by a consultant interpreter (Downie, 2020, p. 92) and performed by trained STTI professionals. Outside this inner interpreting network, the consultant interpreter interacts with different entities for the efficient provision of the service, such as users, institutional stakeholders, and the wider public (Skaaden 2021). By means of two case studies based on observation protocols, we will shed light on the service provision and decision-making entities in the extended interpreting network. We hereby elaborate on the parameters which are taken into account when opting for human or AI-based STT services as well as on the challenges and opportunities which arise for STTI professionals in this specific context of accessibility service aimed for successful and inclusive communication.