Introduction
Artificial intelligence (AI), among other Fourth
Industrial Revolution technologies, has seen
rapid adoption rates, with governments lagging
in regulating the technologies to protect society
from potential negative consequences of their
use in areas of high public sensitivity, such as
healthcare. There is a strong need for alternative
“soft regulation” mechanisms to bridge the ensuing
governance gap. The World Economic Forum
has taken a systematic approach to developing
such soft regulation mechanisms in the form of
frameworks and toolkits for the responsible use of
AI in several applications, including procurement,
human resources and the use of facial recognition.
Chatbots RESET is one such framework.
Working with a multistakeholder project community,
the World Economic Forum has developed and
published Chatbots RESET, a framework for the
responsible use of chatbots in healthcare. The
development of the framework is in response to the
increasing adoption of AI – especially chatbots – in
healthcare settings worldwide. While the use of
chatbots in various healthcare situations was being
experimented with before, the COVID-19 pandemic
created urgency in disseminating information
quickly to the public. Chatbots were widely
deployed for this purpose.
The framework contains a set of AI and healthcare
principles and a set of actions for each of these
principles. The actions provide recommendations
for different players in the system using chatbots in
healthcare: developers (who create the technology),
providers (who use them in healthcare settings)
and regulators (who are responsible for protecting
the public).
Chatbots RESET is a multistakeholder
framework for the responsible use of
chatbots in healthcare.
An important way to quickly validate the Chatbots
RESET framework is to pilot the framework in
real-life applications. Each organization piloting the
framework selected a subset of the framework
principles and tested them in their application
context. Working with four partners, the Forum
piloted the framework from January to April 2021 and
published the results
1
in November 2021. Since then,
the Forum has been working with global healthcare
company Babylon and its Rwandan entity Babyl to
implement key principles from the framework for an
AI triage tool. This tool is part of Babyl’s partnership
with the Government of Rwanda to provide an
integrated digital-first health service that leverages the
country’s universal health insurance scheme, giving
millions of patients access to healthcare in Rwanda.
Piloting
Ethical use of AI is our responsibility – most of all in areas like healthcare.
Kay Firth-Butterfield, Head of Artificial Intelligence and Machine Learning,
World Economic Forum
Chatbots RESET Framework: Rwanda Artificial Intelligence (AI) Triage Pilot 5