- Veröffentlicht: 21. Januar 2019
Smart Dubai has released guidelines for organisations for the ethical use of AI.
The ‘Ethical AI Toolkit' has been created to encourage organisations to develop AI services that prioritize fairness, transparency and accountability.
The guidelines have been developed by Smart Dubai because the smart government entity says there is a need for initial guidelines while regulations are being developed, so that AI development can go ahead without stifling innovation.
"Our vision is for Dubai to excel in the use of technology to maximise human benefit and happiness, as well as to be a global technology standard-setter. Artificial Intelligence plays an integral role in all of this. And with the use of AI growing exponentially across the globe, the ethical dimension of this nascent but rapidly proliferating technology is an increasing topic of discussion on the international stage," said Her Excellency Dr. Aisha bint Butti bin Bishr, Director General of the Smart Dubai Office.
"There is an understanding by governments, NGOs and the private sector that AI regulation is needed, but that the field is not yet mature enough to devise fixed rules to govern it. However, organisations still require guidance and regulators still need to begin to learn how to oversee this emerging technology, but without creating restrictions that could stifle innovation. Smart Dubai's Ethical AI Toolkit aims to provide advice in this area for all those involved in the AI sector," she added.
As part of the toolkit, Smart Dubai has also launched the world's first city-government endorsed AI Ethics Self-Assessment Tool. The AI Ethics Self-Assessment Tool is built to enable AI developers and operators evaluate the ethical level of their AI system, if implemented using Smart Dubai's AI Ethical Principles and Guidelines.
Smart Dubai's Ethical AI Toolkit was created using a benchmarking exercise and a consultation approach. Government sector entities, such as the Telecommunications Regulatory Authority, Dubai Electronic Security Centre, Dubai Health Authority, the Roads and Transport Authority, Dubai Municipality, Dubai Electricity and Water Authority and Dubai Land Department, were consulted during the initial feedback gathering, as were private sector companies including Microsoft, IBM, Google, Etisalat and PWC.
The smart city authority says it welcomes feedback from the AI community, which, along with its research, will help the Ethical AI Toolkit to evolve with AI technology advancements. The office is also establishing an Advisory Board, comprising leading AI and ethics experts from the private and public sectors, who will review the guidelines and help make continuous ongoing improvements to them.
The office highlighted that it would like to see the Ethical AI Toolkit evolve into a universal, practical and applicable framework that informs ethical requirements for AI design and use and one that offers tangible suggestions to help stakeholders adhere to the ethics principle.
"By fusing data and innovation we're preparing Dubai to become the AI city of the future. Artificial Intelligence will streamline day-to-day work life by providing fast and easy access to a wealth of data-driven information. Its consequences will be far-reaching and will impact every area of life, so creating guidelines for AI operatives is essential to provide an ethical underpinning to this evolution. Our aim is to offer unified guidance that is continuously improved in collaboration with our communities, with the eventual goal being to reach widespread agreement and adoption of commonly-agreed policies to inform the ethical use of AI, not just in Dubai but around the world," said His Excellency Younus Al Nasser, Assistant Director General, Smart Dubai and CEO, Smart Dubai Data.
Smart Dubai's Ethical AI Toolkit addresses some of the key issues around establishing regulatory principles relating to AI. These include the rapid evolution of the AI landscape that is leading to a fragmented approach to ethics, with each company dealing with ethical issues in their own way. They also intend to clear the ambiguity around what constitutes ethics in AI, as it is thought that ambiguity could supress innovation through entities holding back on research because they are unsure of future government actions. The toolkit also aims to improve trust in AI systems, with confidence in them cemented by the public being able to see that companies are following the new advice that is published transparently online.
Autor(en)/Author(s): Mark Sutton
Quelle/Source: Arabian Industry, 14.01.2019