Bridging Theory, Policy, and Practice in the Age of AI
The symposium is supported by a grant from the Research Grants Council of the Hong Kong Special Administrative Region, China (Project Reference No.: UGC/IDS(R)14/23).
Register Now
The complex interplay of Globalisation, Internationalisation, Localisation, and Translation (GILT) constitutes the fundamental architecture of contemporary global interaction. This symposium is conceived as a dynamic incubator for pioneering research that interrogates the evolving role of translation—not as a solitary act of linguistic transfer, but as a core socio-technological practice embedded within broader geopolitical, cultural, and economic systems.
Structured around three interconnected pillars—theory, policy, and practice—the symposium brings together academic researchers, legal practitioners, industry professionals, and creative technologists to examine how AI and digital transformation are reshaping the translation landscape. From healthcare communication and creative AI applications to legal frameworks, business ethics, and cultural preservation, we explore translation as it operates across professional domains, regulatory environments, and creative industries.
We seek to foster rigorous dialogue that is both theoretically ambitious and empirically grounded, creating space for doctoral candidates, postdoctoral researchers, early-career faculty, and industry practitioners to collaboratively shape the future of GILT research. By bridging academic enquiry with real-world application, this symposium positions translation studies at the intersection of technological innovation, cultural mediation, and global policy formation.
This session explores critical intersections where translation meets law, healthcare, and emerging technologies. From legal frameworks governing AI to the ethical complexities of medical translation, we examine how translation operates in high-stakes professional contexts where precision, ethics, and human lives are at stake.
Abstract: The Tower of Babel's collapse symbolizes humanity's primal disconnection—ambitious unity shattered by linguistic chaos, mirroring today's global barriers in communication, culture, and commerce. Could AI mend this ancient rift, reconnecting the world through GILT: Globalisation, Internationalisation, Localisation, and Translation? This query strikes at GILT's core, a framework born in the late 1990s to propel businesses worldwide via translation foundations, cultural adaptations, technical preparations, and expansive reach. After 35 years, AI dominates, with the global language services market valued at USD 71-76 billion in late 2025, surging toward $72-90 billion amid e-commerce booms, legal demands, and tech needs. Hong Kong exemplifies this, its $134.35 million NLP-driven sector fueling finance, arbitration, and trade in Cantonese, English, and Chinese—yet judicial inefficiencies persist, with bilingual court transcripts riddled by errors, delays averaging 397 days, and costs soaring.
AI offers Babel-solving potential in real-time transcription and interpretation. Witness Shenzhen's International Arbitration Court: hybrid hearings with live AI transcripts edited onscreen, delivered instantly at 10% of Hong Kong's expense—realizing Richard Susskind's efficient justice vision ahead of Western adoption. Hong Kong courts, burdened by oral proceedings and dual-language hurdles, could leap forward, slashing backlogs and. Yet public sector inertia—bureaucracy, job fears, risk aversion—mirrors private hesitations, questioning if AI fully bridges divides without human oversight.
Global risks reshape ambitions: tariffs, nationalism, and compliance deter full internationalisation, prompting SMEs to fortify domestic markets like China's middle class. This refocuses GILT—less broad globalisation, more hyper-localisation and precise regional translation. MNCs hybridize brilliantly: McDonald's crafts McAloo Tikkis for India while globalizing branding; IKEA tailors China stores; Starbucks brews sakura lattes in Japan; Unilever empowers local subsidiaries. AI accelerates these, blending 90% accurate neural translation with human nuance for legal and cultural finesse.
AI powered this analysis—swift market insights, trend dissections—yet true connection demands Xue Wen: diligent study fused with inquiry, empathy, and ethics. Technology processes data; humanity poses profound questions, like reversing Babel. In our tech revolution, preserve virtues amid flux—AI connects linguistically, but wisdom forges unity. GILT, amplified by AI, holds the key: not just solving Babel's fall, but building a hyper-connected, culturally rich world. Let's embrace this hybrid path forward.
Abstract: Medical translation is one of the most demanding areas of professional practice, combining linguistic precision with domain-specific knowledge and high ethical responsibility. Drawing on real-life experience, the session will discuss ways to build subject-matter understanding—even without a formal medical background—and the reasons why specialising in this field remains professionally rewarding, especially in a market increasingly shaped by AI and post-editing workflows.
Attention will also be given to the wide spectrum of medical text types, such as medical reports, clinical trial documentation, technical manuals, and educational materials, each with its own communicative purpose, intended audience, and typical source-text quality.
A further focus will be placed on the role of machine translation and large language models. Although medical texts are high-risk and unlikely to be fully entrusted to automation in the near future, AI already performs convincingly and can both support and mislead translators. The session will highlight strategies for responsible use of AI tools, the heightened difficulty of detecting errors during post-editing, and the importance of critical verification.
Overall, the session aims to provide realistic, experience-based understanding of the opportunities and challenges in medical translation today.
This panel addresses the ontological core of translation studies, examining the profound redefinition of the practice through technological advancement. How do we prepare the next generation of translators for an AI-augmented profession? What ethical frameworks guide automation? How do we ensure equity and inclusion in an increasingly technologised landscape?
Abstract: The rapid development of machine translation and generative AI has transformed professional translation workflows, but translator-training institutions are struggling to keep pace with these technological and market shifts. This presentation examines the current state of MT and genAI integration in university-level translator education, drawing on mixed-method research conducted at Czech universities and complemented by a case study in Spain.
The session focuses on systemic and institutional obstacles that hinder meaningful inclusion of these technologies in translator training. These include insufficient teacher expertise, reluctance or uncertainty amongst translators, limited resources, and low curricular flexibility. Empirical findings show a mismatch between the realities of translation practice and the training students receive: MT/GenAI use is often discouraged or prohibited without providing students with structured guidance, students lack opportunities to develop high-quality post-editing skills, and although most teachers agree that MT and GenAI deserve a more prominent place in training, they rarely implement them in their courses, even when institutional conditions would allow it.
The session will offer suggestions on how universities and teachers might approach MT and AI more systematically, including clearer communication of technology-use policies and more structured integration of MT and GenAI into existing translation courses to better prepare graduates for today's professional landscape.
Abstract: The use of generative artificial intelligence (GenAI) has provided a much wider array of possibilities for the automation of translation workflows. Such automation has presented considerable benefits, yet at the same time, also ethical concerns. Researchers discussing these prospective ethical impacts have covered a breadth of issues, such as privacy, autonomy, responsibility (e.g., Moorkens, 2022), fairness (e.g., Nurminen & Koponen, 2020), transparency (e.g., Li, 2025), sustainability (both professional and environmental; e.g., Moorkens et al., 2024), and intellectual property (e.g., Li, 2023), which relate to the impacts of GenAI on both translation and translators, as well as its broader effects on the environment and society at large. Yet, these discussions are based on the current capabilities and use cases of GenAI in the context of translation, as outlined in Rivas Ginel and Moorkens (2025). As technology evolves rapidly and often becomes more powerful and enables more use cases in a short time, ethical deliberation on such technology must be based on what may happen in the future. In this paper, therefore, I employ futures studies methods, more specifically, scenario planning, which constructs rich narratives about both the different scenarios and the underlying factors and assumptions that determine them, to envision six potential translation workflows that may be automated with GenAI in the future. For each workflow, I postulate several levels of automation. Whilst previous discussions about the ethics of translation automation have tended to highlight only ethical issues with good or bad visions, my scenarios are multifaceted and consider the value conflicts and moral dilemmas that are likely to be at play in the forecasted futures of automation with GenAI, or even more advanced technology.
Abstract: Thailand has promised to promote equality and inclusion through the ASEAN framework, yet its translational spaces—both physical and digital—remain insufficiently developed to make it a genuine accessibility to life-essential information for all migrant workers. Previous research shows that Thailand's translation/interpreting structures are ineffective, inadequately funded, and too dependent on informal networks, such as volunteer interpreters and community support that, in fact, may not always available (Phanthaphoommee et al., in press; Phanthaphoommee, 2022; Techawongstien & Phanthaphoommee, 2022; Pocock et al., 2020; Min et al., 2023). Schemes like Migrant Health Workers and Migrant Health Volunteers are only part of the meagre governmental response to their linguistic needs. They only work in certain public health settings and are without legal or financial stability (Kosiyaporn et al., 2022; Srichampa et al., 2019).
Drawing on Cronin and Simon's (2014) concept of the translational space, particularly digital connectivity, and its later reformulations in Song's (2024) translational landscape, this study examines whether the "online spaces for translation" of Thai governmental agencies in charge of migrant workers promote or impede true inclusivity for the marginalised. Through a case study of official Thai governmental agencies' websites and digital communication portals designed for migrant workers, I show how the translation (or lack thereof) in these online platforms can be regarded as a social and political act. Preliminary findings suggest a "reluctant positionality" of the Thai state. That is, translation is deployed to perform international compliance and conformity rather than to assiduously cultivate intercultural reciprocity and recognition. This paradox exposes an aperture between Thailand's commitments to ASEAN equality principles, such as the ASEAN Consensus on the Protection and Promotion of the Rights of Migrant Workers (Association of Southeast Asian Nations, 2018) and its limited purposeful implementation of language policies.
By situating Thailand's translation policies as "bureaucratic translation" (e.g., Coutin & Fortin, 2023; Cortés, 2029), I argue that digital connectivity between the powerful and the powerless via governmental online spaces have yet to become genuinely inclusive zones for all migrant workers. They reveal, instead, how linguistic accessibility is constrained by sociopolitical pragmatism, budget limitations, rigid hierarchies in a monolingual society, and meanings of the citizen vs the Other.
Cultural products and narratives do not circulate freely; their movement is mediated through translation, which acts as a critical site of power, adaptation, and resistance. This session examines how translation connects diverse worlds—from global business networks to social justice initiatives to the preservation of digital cultural heritage.
Abstract: Business translation is increasingly relevant in both professional practice and academic research. Yet, scientific production in this field reflects diverse dynamics that extend beyond disciplinary boundaries. Drawing on a corpus of more than 3,000 publications compiled within the TRADECO project (Business Translation and Research. A Bibliometric-Based Study. MCIN/AEI/10.13039/501100011033), the study applies a set of bibliometric indicators to map global production patterns and to explore potential signs of a slowdown in research activity, tendencies towards locally oriented research, and the extent to which scholarly output aligns with professional translation practice. Amongst other aspects, the findings reveal notable contrasts across regions. Spain and China emerge as leading contributors. Linguistic patterns also differ: whilst Western contexts display a certain degree of multilingualism, English and French dominate in Arabophone regions, and Chinese is predominant in East Asia. Keyword analysis highlights global convergences around themes such as foreign trade and corporate communication, alongside divergences shaped by local research traditions and professional environments. Author and journal profiles further suggest that the Americas tend to concentrate on practical and applied translation issues. The talk aims to reflect on the sociological, cultural, and technological factors that may be shaping these patterns and to encourage new generations of scholars and practitioners to build bridges across cultures, disciplines, and research traditions.
Abstract: As generative AI-powered translation tools become increasingly embedded in UK higher education, they are reshaping how international students who use English as an additional language and academic staff engage with academic communication in learning and teaching practices. These tools function not only as forms of linguistic support but also as mediators of disciplinary access, academic voice, and multilingual identity within English-dominant academic environments. However, their rapid adoption has introduced uncertainty and mistrust, particularly where understandings of GenAI-powered translation practices diverge between students and staff and where concerns arise around academic integrity, learning, and originality. This qualitative study will investigate how international students who use English as an additional language at UK universities draw on these tools to interpret disciplinary content, articulate complex ideas, and participate in academic discourse, and how academic staff perceive, interpret, and respond to such practices. The findings will inform practical recommendations for the ethical and equitable integration of GenAI-powered translation tools, highlighting the importance of developing AI competency amongst both students and staff. This includes fostering a shared understanding of ethical academic AI use, strengthening skills in critically evaluating AI-generated translations, and increasing confidence in communicating expectations and boundaries between students and staff concerning AI in academic work.
Abstract: In a world increasingly troubled by rising inequality, geopolitical tensions and tribalism, there is an urgent and growing call for initiatives that promote social good to counter these negative forces. There is, however, a lack of comprehensive empirical studies that examine the role of translation in the three key domains of social good – diversity and inclusion, environmental justice and sustainability, and peace and collaboration (Barak 2020). Much of the existing research falls short in exploring the potential of AI-based translation for social good, notwithstanding AI's growing influence.
This paper seeks to investigate the role of translation across all three domains of social good and argues that advancements in AI can strengthen, rather than undermine, the impact of translation on fostering positive changes and that human translators remain central to these efforts. Singapore provides a compelling case study as its bilingual policy necessitates translation services, and its growing prosperity paves the way for an emphasis on social good.
The study begins with a multidisciplinary literature review, followed by a two-phase qualitative data analysis focusing on public communication within healthcare, social and legal services. First, the study identifies gaps that contravene the principles of inclusivity, sustainability and harmony by analysing specific examples related to translation. Whilst it elucidates translation's symbolic role which reflects Singapore's commitment to linguistic and cultural diversity, significant gaps remain in safeguarding the linguistic rights of non-English-speaking individuals. Second, the study evaluates the potentials and risks of deploying different models of AI-based translation to address the deficiencies identified in the earlier phase. Although AI enhances efficiency and capacity, it poses serious risks including homogenisation, discrimination and misinformation, further exacerbating ethical and socio-political issues that are intricately linked to the notion of social good.
The case study of Singapore reveals the transformative potential of AI-based translation in advancing social good, contingent on the involvement of human translators, whose contributions are essential to promoting linguistic equity, environmental sustainability and cross-sector collaboration.
Abstract: The integration of artificial intelligence technologies and non-Western music presents a unique opportunity for computational methods and music representation through culturally informed insights. Current AI music systems suffer from documented structural biases that systematically marginalise non-Western musical traditions. In this talk, my objective is to present and discuss the field of AI and Music, with a particular focus on non-Western music.
Standard MIR pipelines embed datasets and models in Western-centric assumptions about elements of music, which fail to capture the nuances of non-Western music traditions. Non-Western music traditions (e.g., Chinese, Indian, and African music practices) are rooted in oral transmission, adaptive pedagogy, and multimodal channels—such as vocalisation, rhythm, dance, stories, gesture, and embodied enactment—that can offer valuable insights for multimodal AI systems.
Developing a framework which involves communities actively annotating and building datasets that reflect their own cultures, treating oral traditions as dynamic systems of knowledge sharing, may minimise bias and address data shortages, particularly in resource-constrained settings. Yes, an interdisciplinary and domain-specific approach holds promise for advancing future possibilities in the field of AI and music.