Overview
-
Artificial intelligence (AI) is the capability of computational systems to perform tasks typically associated with human intelligence, such as learning, reasoning, problem-solving, perception, and decision-making. It is a field of research in engineering, mathematics, and computer science that enable machines to perceive their environment and use machine
learning to take actions that maximise their chances of achieving defined goals. AI comprises a wide range of technologies and disciplines, including vision, perception, speech and dialogue, decision-making and planning, problem-solving, robotics and other applications that enable self-learning.
- AI offers the potential to advance solutions for people and the planet as well as contributing to economic growth. At the same time, its adoption raises significant social, economic and ethical considerations. AI developments are closely linked to data governance, including data protection, privacy and related policy frameworks.
- Advances in AI are closely connected to advances in data processing, computing power and digital infrastructure. The impact of AI also varies with different countries' existing economic structures and levels of digitalization, particularly in sectors such as health, agriculture and manufacturing.
- The International Telecommunication Union (ITU) works on artificial intelligence from a broad perspective, with activities covering
AI solutions and use cases, technical standards, skills development and capacity building. This work contributes to an enabling environment for AI innovation and adoption, while supporting sustainable development and inclusive access to its benefits.
What is AI?
AI can be defined in several ways:
-
By technologies or techniques, such as neural networks used in machine translation.
- By purpose, such as facial recognition or image recognition.
- By function, including understanding language, recognizing images, solving problems and learning.
- By systems or agents, such as robots or autonomous vehicles.
-
AI and machine learning (ML) have advanced rapidly with growth in computing power, processing speed and advanced data management.
Opportunities
AI offers opportunities to improve scale, speed and accuracy across many sectors. Combined with statistics and big data, AI can support more predictive, proactive and evidence-based approaches to policymaking, regulation and business operations.
AI technologies are already widely used in:
- Online search and entertainment.
- Social media.
- Autonomous and assisted driving systems.
- Visual recognition.
- Translation and publishing tools.
- Smart assistants and voice technologies.
- Speech-to-text applications.
AI can help accelerate sustainable social and economic development in key areas such as:
-
Healthcare: Remote health services, diagnosis support, predictive analytics and improved clinical decision-making.
-
Education: Personalized learning, identification of learning difficulties and adaptive educational tools.
-
Finance: Automation of routine tasks, data-driven financial services and support for financial inclusion.
-
Manufacturing: Automation, fifth-generation (5G) mobile technologies, the Internet of Things (IoT), robotics and more efficient production systems.
-
Transport: Autonomous vehicles, assisted driving systems and traffic management.
-
Agriculture: Farm management, crop monitoring, predictive analytics, disease detection and optimization of resource use.
In addition, AI platforms can support public engagement and information, with AI-based translation tools facilitating multilingual communications, content distribution and customer-service applications.
Challenges
AI policy and regulatory frameworks continue to evolve, as governments and regulators worldwide strive to balance innovation, safety, accountability and public trust. The approaches taken differ across jurisdictions, reflecting differing national priorities and capacities.
Questions continue to arise about:
- The accuracy and reliability of AI systems.
- Accountability and responsibility for AI decisions.
- Intended and unintended uses of AI applications.
- Methods used to collect and process training data.
- Data ownership and intellectual property.
- Data privacy and security.
Concerns have also been raised regarding the quality and representativeness of datasets used to train AI systems. Researchers continue to examine issues relating to bias, including racial, gender, disability and socioeconomic bias.
Additional challenges include:
-
Trust and transparency: AI models, particularly deep-learning systems, may be difficult to interpret or explain, raising questions about transparency and accountability.
-
Bias: AI systems may reinforce existing inequalities when training data are incomplete, unrepresentative or poorly labelled.
-
Data availability and ownership: Access to data remains difficult, while balancing innovation with confidentiality and ownership rights.
-
Data privacy and security: AI systems rely on large volumes of data, increasing the importance of privacy-preserving technologies and cybersecurity protections.
-
Limited know-how: A shortage of expertise in the responsible and ethical deployment of AI necessitates broader education and interdisciplinary participation.
-
Uneven access:
AI research and deployment require significant computing resources and access to data, creating the risk of widening digital and innovation divides.
ITU's contribution
AI for Good
The AI for Good platform focuses on practical applications of AI that support progress towards the SDGs. The initiative brings together stakeholders from across the United Nations system, governments, industry, academia and civil society.
Resources:
Policy and regulatory support
ITU monitors developments in national AI strategies and regulatory approaches through its regulatory and market monitoring activities.
ITU highlights the importance of considering data protection, privacy, IoT deployments, sensor networks and 5G infrastructure when developing national AI strategies.
Resource:
Technical standards
International standards help address real and perceived risks by improving the predictability, reliability and efficiency of AI and ML systems.
ITU-R and ITU-T study groups are examining AI applications in radiocommunications, broadcasting, telecommunications networks, digital health, multimedia, autonomous driving, environmental sustainability and network management.
Examples include:
- ITU-R studies relating to spectrum management and broadcasting applications of AI.
- ITU-T Recommendations supporting ML in 5G and future networks.
- Standards and studies addressing multimedia quality assessment, network analytics, datacentre management, environmental efficiency and cable networks.
- The ITU AI/ML in 5G Challenge
- ITU-T focus Group on Environmental Efficiency for AI and other Emerging Technologies
- ITU-T Focus Group on AI for Health.
- ITU-T Focus Group on AI for Autonomous and Assisted Driving.
- AI standards exchange database
- Resilience to natural hazards through AI
- AI Native for telecom networks
- Embodied AI for Multimedia Technologies
AI skills and capacity development
ITU supports the dissemination of knowledge on AI applications for development through publications, pilot projects and partnerships.
ITU works on a range of
partnerships and global initiatives to promote responsible, sustainable AI.
For example, the
AI Skills Coalition brings together over 80 partners to deliver online and in-person AI training.
Other key AI-related partnerships include:
-
AI and Multimedia Authenticity Standards Collaboration
-
Global Initiative on AI for Health
-
Global Initiative on AI for (Agri-) Food Systems
Further resources
Last update: October 2026
