ARTIFICIAL INTERPRETING
this real-time communication takes place in a huge
range of communicative settings – from low-stakes
tourism inquiries to high-stakes medical, legal and
diplomatic communication. When establishing best
practices for interpreting, this continuum will need to
be considered.
How will SAFE-AI's work impact the field in the US, and
how do you see the organisation coordinating with
other bodies towards a unified global approach?
The United States, unlike other countries, does have
more of a governance structure in place to safeguard the
rollout of AI in interpreting, especially when it comes to
settings that must comply with broad and established
language access requirements, such as for federally-
funded healthcare, education and public services. Given
the liability risks and potential for serious harm if AI
applications were applied indiscriminately, the rollout of
AI interpreting to replace human interpreters is likely to
be slower in these areas.
The SAFE-AI Task Force represents a wide spread of
stakeholders interested in an effective and intentional
integration of AI into interpreted communication. This
broad representation will help us in our mission to
establish, disseminate and promote industry-wide
guidelines and best practices for the responsible
adoption of AI in interpreting.
The group is in its initial stages but it has a
governance and committee structure designed to allow
for continued research and engagement with an
increasing number of groups after we publish the results
of the perception surveys. Our goal is to collaborate
effectively with major efforts to influence how AI impacts
the interpreting industry. We are taking this matter
seriously and encourage you to do the same. Educate
yourselves and all other stakeholders about these
technological developments and their consequences,
and hopefully do it with more eloquence and accuracy
than ChatGPT3.5!
Notes
1 The US folk hero who raised the alarm as the British forces
approached in 1775; https://cutt.ly/PRev
2 Henceforth referred to as ChatGPT
3 https://cutt.ly/zdnetAI
4 https://cutt.ly/uncannyvalley
5 https://cutt.ly/TechTarget
6 https://safeaitf.org
12 The Linguist Vol/63 No/1
thelinguist.uberflip.com
©
PEXELS
SAFEGUARDS
Stakeholders are
working on a
framework to deal with
issues such as bias,
data breaches and
civil rights protection
