2022 has been a very important year for the development of Artificial Intelligence. Algorithms like Stable Diffusion or Dall-E have shown that digital art can be produced simply by writing what we imagine in a text box; OpenAI has just presented an intelligent chat that already competes with Google searches and more and more companies are making business decisions based on intelligent recommendations.
However, it is also true that AI poses a series of important ethical problems, such as its lack of responsibility when making decisions, lack of transparency (how algorithms are programmed), possible discrimination of people based on external criteria, respect for the privacy and intimacy of individuals and in the not too distant future, the possible destruction of jobs.
Faced with this situation, Fujitsu made its promise last year to offer secure and transparent AI technology. That intention was reflected in the Fujitsu Group AI Commitment. Also, I create a Office of AI Ethics and Governance, in order to develop and implement robust policies for AI ethics and ensure their effectiveness.
This effort has been translated into practice, having developed in 2022 a set of tools and guidelines for developers, which serve as a guide to assess the possible ethical implications that may occur in the algorithm in which they work.
This kit is made up of case studies and reference materials, including a whitepaper, a procedure manual, and a method for clarifying requirements in artificial intelligence ethics guidelines written in natural language and is based on a model interpretation of the guidelines published by the European Commission on this matter. The intention is to avoid misunderstandings and other problems due to differences in the interpretation of the descriptions of the guidelines, offering the possibility of detect and prevent ethical failures in the early stages of the development process.
These resources are available for free, with the aim of promoting the safe deployment of artificial intelligence systems in society and for all this, stands with the TPNET Award 2022 in the category “Innovation in AI”.