This was the most heavily used set of boards, spanning both a short opening-question board and a longer table-by-table board. The content is grouped by sub-topic below rather than by table, since similar points recurred across several tables.
What ethics means to delegates
In response to the opening question ("what does ethics mean to you as a language professional?"), delegates most often described ethics in terms of:
- Honesty, integrity and doing the right thing
- Accuracy and faithfulness to the source text
- Confidentiality
- Fair pay for good work
- Professionalism and maintaining high standards
- Transparency with clients and colleagues
- Being honest about the limits of one's own expertise, and declining work outside it
- Considering the wider impact of the profession's work on people and the planet
A smaller number of delegates linked ethics explicitly to sustainability of the profession as a whole, or to the environmental impact of AI and data centres.
AI: accountability, transparency and control
- A recurring question was who benefits from AI use, who is held accountable when it goes wrong, and how honest AI providers are about what they are selling.
- Delegates discussed the pressure to use AI even when a client has not asked for it, particularly under time pressure — described by some as "bending the rules".
- Several delegates raised concerns about the origins of AI training data, describing it as built without compensation to the original creators of the text.
- One delegate compared post-editing work to being a "reverse centaur": professionals are given full accountability for catching every AI error, regardless of the quality of the AI output, rather than AI genuinely assisting the human.
- Delegates noted a loss of control paired with maximum responsibility, and suggested a shift toward humans being "in the driving seat" rather than simply "in the loop".
- A question was raised about whether translators or translator associations should consider building their own tools, rather than relying on tools built by others.
- It was noted that the French and Spanish translation sectors have published statements against AI, and that a "slow translation manifesto" was seen by some delegates as a step in the right direction.
Professional duty and difficult decisions
- Several delegates returned to the principle that a translator's primary duty is to the client, and that professionals should decline work they cannot accept on ethical grounds (for example, in defence or fields the delegate is not equipped for), rather than complete it and stay silent.
- Delegates discussed whether it is ethical to work into a second language, noting that this is increasingly common and, in the view of some, increasingly accepted.
- The question of whether to tell a new client that AI is not used in their work was raised, without a settled answer.
- Some delegates linked ethical strain to freelance working conditions specifically: an individual freelancer carries a heavier mental load in trying to "do the right thing" than someone working inside an organisation with dedicated compliance functions.
Sustainability, fairness and the profession's future
- Delegates drew a parallel with the closure of the UK coal mining industry, suggesting that AI's impact on the translation profession could be similarly disruptive if the transition is not well managed, and expressed concern that government policy currently favours rapid AI adoption.
- Questions were raised about whether it remains ethical to train or encourage new entrants into the profession, given the pressure on job opportunities.
- Some delegates questioned whether ethics is something only better-resourced professionals can afford, noting that translators who cannot afford conference attendance may also be less able to afford acting ethically.
- Rates and unfair competition were raised, alongside a comment that the profession "is not a charity".
- Delegates discussed carbon offsetting and whether AI use is always less sustainable than the alternative, noting that this may depend on the specific comparison being made (for example, digital tools versus paper-based alternatives, or travelling to an assignment versus using remote technology).
Sector-specific and vulnerable-context dilemmas
- Delegates discussed the use of untrained or volunteer interpreters by NGOs in crisis and conflict settings, weighing the benefit of expanded access to language support against the lack of quality control and unclear scalability.
- Concerns were raised about the safety of interpreters working in conflict zones, with a link shared to Red T, an organisation supporting interpreters and translators in high-risk settings.
- In publicly funded settings such as court interpreting, delegates felt AI is not yet good enough (citing issues such as transcription hallucinations), and discussed the value of complaints in driving service improvements, alongside a concern that government bodies may push for AI-only provision without a human in the loop.
A comparison was raised between the ethics of directing humanitarian budgets toward human translation versus other priorities such as global health programmes, without a clear resolution.