One year ago, our Chief Executive Officer, Warren East, announced at London Tech Week 2020 that, after a humbling extensive peer review of our AI ethics and trustworthiness toolkit, we had made two breakthroughs and that we would publish to invite use, critique and improvements.
One of our objectives was - and continues to be - to contribute to building public trust in AI so that we can further unlock human potential and accelerate greater industrial, commercial and societal endeavours.
The Aletheia FrameworkTM was published on Rolls-Royce.com in December 2020 under Creative Commons licence. Even though it’s the start of a long journey, it has been downloaded nearly 2,000 times; cited in recommendations from the Institute for Ethical AI in Education; and been the subject of a BBC documentary.
This is in addition to collaborations with AI practitioners outside of our sector to test the applicability of The Aletheia Framework in non-industrial contexts.
Our activity has been so well received that we’ve decided to publish The Aletheia Framework 2.0 later this year, incorporating the critique we’ve had, as well as new module created by our assurance expert, Lee Glazier. This continues the step-by-step approach by offering a process for assessing, identifying and mitigating bias risk in AI requirements, algorithms and data sets that are used in the development and use of AIs.
As a global power group, Rolls-Royce is focused on leading the transition to net zero and we believe the world cannot get to net zero without artificial intelligence. There is lots still to do but The Aletheia Framework is now a key component of our sustainability journey.
But the purpose of The Aletheia Framework is much bigger than just improving Rolls-Royce and the products and services we can offer to our customers. Our initial peer review process, coupled and this year's extensive engagement activities, strongly indicate that we are at the leading edge of operationalising ethical considerations and assuring trustworthiness.
The Aletheia Framework clearly doesn’t have all the answers and the critique we have received of The Aletheia Framework is helping us strengthen it. So, we’re really excited about sharing the updated version in December this year and continuing our engagement activities outside our sector to contribute what we can to the global discussion on AI.