Introduction
Artificial Intelligence (AI) is increasingly present in our everyday lives, whether through smart assistants, crypto currencies, recommendations when we search the web, chatbots, etc. AI’s fields of application are manifold and rapidly continue to expand as the digital transformation accelerates due to the COVID-19 pandemic and technological breakthroughs.
AI technologies have the capability to support sustainable development and positively transform people’s lives, yet they may also have unanticipated effects on societies, economies and the environment.
UNECE is assessing the use of AI in its areas of work to help member States harness AI in a manner that is based on the United Nations Charter and the Universal Declaration of Human Rights.
Large Language Models for Official Statistics
Established in 2010, the UNECE High Level Group on Modernisation of Official Statistics (HLG-MOS) has been at the forefront of modernisation initiatives in the field of official statistics. These initiatives include innovative areas such as big data, synthetic data and machine learning – AI system developed based on data.
In 2023, the international collaboration efforts of HLG-MOS on the topic of AI have concentrated around the Large Language model (LLMs) – advanced AI systems that sparked a substantial public interest with the release of ChatGPT in late 2022. Based on the extensive training on vast data sets, LLMs are capable of understanding and generating texts that cannot always easily be differentiated from texts produced by a human being. This exceptional capability in natural language processing tasks offers an immense potential for statistical organizations to enhance the quality of their services to society – provision of statistics and data services that are foundational for policy-makers, business and citizens alike.
To establish a common understanding of LLM’s potential within the statistical community, a recently published white paper provides examples of implementation from various statistical organizations (e.g., code translation, report generation, natural language interface for database), which showcase the capabilities, challenges, and risks of LLM (e.g., privacy concerns, hallucination, ethical issues, governance, alignment with Fundamental Principles of Official Statistics).
Developing the regulatory framework for autonomous vehicles
The motor industry’s 125-year history is an impressive succession of innovation. Today, the industry faces the challenging task to massively electrify vehicles over the next 5 to 10 years, at least in developed markets.
At the same time, with the rise and promises associated to automated and autonomous driving, we are standing before what will be the biggest quantum leap forward in automotive technology in history.
Autonomous vehicles have the potential to improve the life of billions of people and fundamentally change how road transportation works. This will transform the global automotive industry, which employs some 50 million people worldwide and represents a turnover of almost $2 trillion per year.
Yet, autonomous vehicles raise as many questions as they offer potential benefits. These include liability, insurance regimes, safety standards, software reliability and cybersecurity, to name just a few.
Providing an appropriate and balanced regulatory answer to these questions is a prerequisite to the mass introduction of these vehicles on the road. UNECE, through the World Forum for Harmonization of Vehicle Regulations (WP.29) and the UN Road Safety Instruments, is the platform where countries from all around the world gather to jointly develop the regulatory frameworks governing motor vehicles and road traffic. It has therefore embarked since 2015 in adapting existing legal instruments and developing new ones to facilitate the gradual introduction of automated and autonomous driving functionalities made possible by the use of AI and other technologies.
In an effort to focus its activities on automation, UNECE established the Working Party on Automated/Autonomous and Connected Vehicles (GRVA) in 2019.
Since its establishment, GRVA has created a global scheme to develop requirements and guidelines for automated and connected vehicles, namely the Framework on Automated/Autonomous and Connected Vehicles (FDAV), which largely guides GRVA’s work. This Framework was drafted by the European Union, China, Japan, and the United States of America and endorsed by the World Forum for Harmonization of Vehicle Regulations and the UNECE Inland Transport Committee. The document defines a safety vision, key safety elements, guidance to the Working Parties of WP.29 as well as a programme of activities. These activities, at the intergovernmental level, form a novel initiative aimed at globally harmonizing automated vehicles regulations and creating a more productive environment for innovation.
Recent developments under the 1958 Agreement include the adoption of amendments to UN Regulation No. 79 (Steering equipment) as well as new UN Regulations Nos.: 157 (Automated Lane Keeping System), 155 (Cyber Security and Cyber Security Management System) and 156 (Software Update and Software Update Management System).
As Advanced Driver Assistance Systems and Automated Driving Systems rely on combinations of technologies including technologies that the general public call Artificial Intelligence, GRVA address the safety of such systems in a technology neutral manner. Given the nature of machine learning and deep learning, for example, GRVA reviewed how industry employ such AI technologies and for which use cases. It drafted relevant definitions for the purpose of its work addressing the automotive sector and it is considering the need to develop AI specific provisions in the form of recommendation or guidelines that would address the specific risks posed by the technology.
AI for Smart Cities
The UN global initiative United for Smart Sustainable Cities (U4SSC) established by UNECE and the International Telecommunication Union (ITU) in 2016 is a global platform for smart cities stakeholders, which advocates for public policies to encourage the use of ICT for sustainable urban development. The initiative is coordinated by ITU, UNECE and UN-Habitat, and supported by 14 other UN agencies and programmes.
The initiative aims to: Generate guidelines, policies and frameworks for the integration of ICTs into urban operations, based on the SDGs, international standards and urban key performance indicators (KPIs); and help streamline smart sustainable cities action plans and establish best practices with feasible targets that urban development stakeholders are encouraged to meet. The topics of the current phase of
U4SSC looks in particular at how AI applications can make cities smarter and more sustainable.
In 2017, the U4SSC stakeholders elaborated a set of KPIs for smart sustainable cities (KPI4SSC) which includes 92 indicators (core and advanced) divided in the 3 dimensions of sustainable development: economy, environment, and society and culture. The indicators are fully aligned with the Sustainable Development Goals (SDGs) and serve as a tool for evidence-based decision making, progress monitoring and achieving the SDGs at the local level. They are being implemented by 50 cities of different sizes and development worldwide.
Cities taking part in UNECE‘s KPI evaluation include: San Marino, Almaty (Kazakhstan), Bishkek (Kyrgyzstan), Nur Sultan (Montenegro), Tblisi (Georgia), Voznesensk (Ukraine), Goris (Armenia), 17 cities in Norway and many other countries
UNECE Housing and Land Management Unit implements a UNDA 12th tranche project on innovative financing for sustainable smart cities and as part of the project, develops smart sustainable cities profiles using the KPIs for SSC for the cities of Grodno (Belarus), Nur-Sultan (Kazakhstan), Bishkek (Kyrgyzstan), Tblisi (Georgia) and Podgorica (Montenegro). Based on the cities profiles recommendations, including for innovative urban infrastructure using AI, UNECE supports with developing bankable projects and connections to the IFIs. More information is at https://unece.org/housing/innovativefinancing-sustainablesmartcities. Also see the publication on “people-smart” cities at https://unece.org/media/Housing-and-Land-Management/press/355189
AI for Energy
Artificial Intelligence and other technologies are inspiring energy suppliers, transmission and distribution companies, and demand sectors (buildings, industry, transport, and other) to establish new business models to generate, deliver and consume energy in a more sustainable way.
UNECE established a Task Force on Digitalization in Energy to offer a platform for cross-industry experts from the energy sector and digital innovation to develop a unified voice on digitalization in energy. Serving as an umbrella for the subsidiary bodies of the Committee on Sustainable Energy to conduct relevant research and assess sectoral opportunities and challenges, the Task Force on Digitalization in Energy:
-
Monitors new and emerging trends in the digitalization in energy domain, that enable advances in connectivity, data, analytics, optimization of the overall energy infrastructure, and can greatly increase overall efficiency of the energy system;
-
Conducts in-depth research into the potential of integrating digital solutions throughout the entire energy system, based on thorough evaluations of challenges and policy obstacles, including notably the socio-economic context, to provide a clear, concise and balanced view to policymakers and other stakeholders;
The aim of the Task Force on Digitalization in Energy is to bridge the gap between academic research, industrial innovations, and policy needs, and to develop a consensus about the approach that should be considered for shaping the future of energy systems.
The Task Force on Digitalization in Energy found that AI and digitalization have the potential to reduce residential and commercial buildings’ energy use by as much as 10% globally by 2040 if applied throughout buildings value chain and life cycle. In particular, applications of AI may help optimize buildings’ orientation for solar heat gains and predict individual power and heat needs, thus increasing overall energy security and maximizing the integration of renewable energy sources.
The Digitalization in Energy also found that AI and digitalization could help achieve energy savings of at least 10-20% in the industrial sector (which consumes around 38% of global final energy and produces 24% of GHG).
Other findings of the Task Force on Digitalization in Energy on Big Data, advanced analytics, machine learning, AI, and related matters, also hold considerable significance for energy sector industries and end uses, and inform the subject-matter discussions at UNECE level and beyond.
Energy systems transformation involves finding a balance between energy security, affordability, and environmental sustainability, which will define their resilience (as defined in UNECE’s publication “Building Resilient Energy Systems”): energy security, affordability and environmental sustainability.
To facilitate decision-making in this complex endeavour, UNECE partners with University of Zürich to develop an AI-powered tool that will use selected and vetted information sources. European Investment Bank, International Atomic Energy Agency, International Energy Agency, International Telecommunication Union, Organization for Security and Co-operation in Europe, World Meteorological Organization, World Bank, and other organizations contribute their knowledge base to support and shape this tool.
AI for Trade
Facilitating International Trade
Artificial Intelligence (AI) is transforming international trade by streamlining information exchange and enabling data-driven decision-making for trade actors along the supply chain. Combined with business-process-oriented automation and more efficient data flow exchanges, AI promises to help address trade barriers, stimulate growth in global e-commerce and inform policy decisions.
The UNECE-hosted United Nations Centre for Trade Facilitation and Electronic Business (UN/CEFACT) plays a key role in this transformation by integrating AI into its data exchange standards. For example, UNECE-developed UN/CEFACT Buy-Ship-Pay JSON-LD Web Vocabulary (2023), in addition to existing XML schema format allows systems to leverage machine-readable data and deep learning algorithms to graphically represent, cross-reference, and comprehend crucial trade contexts. This promises to remove obstacles that have historically slowed global commerce, such as data inconsistency and incompatibility or manual processing delays. If done successfully, this can result in a more connected, efficient, and resilient supply chain well-equipped to meet the demands of the 21st century.
Furthermore, successful integration of AI with automation, blockchain and Internet of Things can further enhance data flow efficiency, optimize supply chains, and improve e-commerce. In this connection a UN/CEFACT Whitepaper on the Use of AI for Trade Facilitation (published by UN/CEFACT experts in February 2023), explores AI’s potential to enhance data collection, processing, and analysis in trade facilitation.
At the same time AI still presents several challenges related to cybersecurity and data privacy risks as well as lack of critical thinking, creativity, and ethical judgment. These risks need to be taken into consideration during the development of regulatory and legal frameworks, as well as during the development of AI applications.
Navigating New Regulatory Challenges
As digital technologies become embedded in products, including products that are traded internationally, this places new demands on the many processes that Quality Infrastructure (QI) oversees, such as standardization, conformity assessment, accreditation, and metrology. QI must adapt to ensure products meet international standards for safety, performance, and interoperability. This is needed to ensure that digitally enhanced products are reliable, secure, and compliant with global regulations, facilitating smoother international trade.
UNECE’s Working Party on Regulatory Cooperation and Standardization Policies (WP.6) has responded to these new demands and developed guidance for regulating products with embedded AI.
Specifically, UNECE’s Common Regulatory Arrangement on the Digital Regulation of Goods and AI, and its related declaration, was approved in April 2023 by UNECE member States. This framework aims to facilitate the seamless cross-border movement of AI-embedded products by harmonizing regulatory requirements. It is expected to enhance international trade efficiency, reduce technical barriers, and foster innovation by providing a standardized approach. Countries adopting this framework can experience smoother trade operations and increased trust in the safety and compliance of AI-embedded goods.
Through these efforts, UNECE is creating a blueprint for a forward-thinking regulatory environment that supports innovation while ensuring product safety and compliance.
AI for Innovation
Increasing the pace of research and innovation: the example of climate resilience
Recent breakthroughs in AI hold the promise of markedly increasing the productivity and hence pace of research, science, technology, and innovation in all areas, including on climate action.
Yet, many applications of AI today have large energy and environmental footprints. As these applications grow exponentially, they risk creating large scale new demands on energy supply and distribution. They also pose significant new challenges for data security, data privacy, data access and ownership, and ensuring that AI is used for the common good.
The need for balancing challenges with opportunities manifests itself in regard to innovation for combatting climate change.
Getting to net zero carbon emissions requires transformative innovation at the level of entire socio-economic systems, across all industries and especially at the level of cities where over 70% of global CO2 emissions is generated. Governments, innovators, investors, businesses, consumers, and civil society need to collaborate to accelerate the development, deployment and scale-up of innovative solutions to climate change, including across borders. Transformative innovation, and AI in particular, will have a role to play in this endeavor; bringing both, opportunities and challenges.
In summer 2024 at UNECE’s Committee on Innovation, Competitiveness and Public-Private Partnerships, UNECE member States explored the opportunities and challenges of AI in addressing climate change. Member States and experts addressed the questions on how AI can accelerate the breakthrough innovations for climate change mitigation and adaptation, how innovation can reduce the footprint of AI itself and what policy makers can do to support this process.
Discussions will continue at the forthcoming 16th session of the Team of Specialists on Innovation and Competitiveness Policy in November 2024 which dedicates its international policy dialogue to the theme “Unleashing the power of innovation for climate action: AI for net zero and climate resilient infrastructure”.
Discussions will: explore the benefits of AI in climate change action - both in mitigation and the transition to net zero and adaptation and resilient infrastructure; highlight the risks and challenges from AI; and identify policy recommendations for governments to steer the use of AI in the direction of a sustainable and equitable future.
The discussion will identify key messages and recommendations to contribute to the next UNECE session in 2025 on the theme “Climate action and resilient infrastructure for a sustainable future”.
AI for Infrastructure Financing
AI offers opportunities in infrastructure public-private partnerships (PPPs): it can bring down the fixed costs of individual projects and thus reduce the minimum efficient scale, thereby enabling large numbers of small-scale projects to be implemented. Beyond the PPP planning process, AI can also have a transformative effect throughout the project lifecycle However, there are also substantial legal, ethical, and technical challenges that arise with the integration of AI. The main risks consist in reliability limitation, the lack of transparency in the algorithms and the data, and the lack of controllability.
The need for balancing challenges with opportunities manifests itself regarding PPPs for climate resilience.
Already in June 2024, UNECE’s Committee on Innovation, Competitiveness PPPs has explored AI’s role in addressing climate change through the “AI for net zero and climate-resilient infrastructure” initiative. Combined these efforts examine how AI can accelerate the transition to net zero, optimize energy use, and develop infrastructure resilient to climate impacts.
Discussions will continue at the forthcoming 8th session of the Working Party on Public-Private Partnerships which will dedicate its policy discussion to the theme “Climate action and SDG acceleration through sustainable Public-Private Partnerships and resilient infrastructure”.
At the session in November 2024, delegates will discuss integrating AI into infrastructure development. Discussions will look at how AI can analyze data to guide decisions on building hospitals, schools, and energy grids. This data-driven approach enhances government planning efficiency, allowing a shift from reactive to proactive development strategies.
Discussions will consider how cutting-edge technologies can be effectively integrated into PPP projects to improve efficiency, transparency, and sustainability; highlight challenges regarding data security, privacy, and ethical considerations when using such technologies; and how AI and other digital tools can be leveraged to improve and streamline PPP projects.
To address these challenges, UNECE is further planning a series of policy briefs on the potential and limitations of AI-enhanced PPPs and infrastructure projects for the SDGs. These policy briefs will enable to measure the efficiencies and ensure that its application in PPPs is ethical and effective.