What is artificial intelligence? It is computer systems that learn from data and then use what they have learned to do tasks we usually link with human thinking, such as understanding speech, translating, summarising and spotting patterns. This page gives a plain overview of AI: what it is, its main types, and the benefits and risks of AI, with reliable sources students can cite if they are writing a report or essay about it.

If you are looking for the core AI terms side by side in English and Arabic, see AI in English and Arabic: the core AI terms side by side. If what worries you is fear of AI, start with our Fears vs Reality series.

What is artificial intelligence? The definition

In plain words. AI is software that is not given a fixed rule for every case. Instead, it learns patterns from many examples and then predicts the most likely answer or result. That is why it can be remarkably useful, and also why it is sometimes confidently wrong (glossary).

The definition international bodies use. The OECD defines an AI system as "a machine-based system that, for explicit or implicit objectives, infers, from the input it receives, how to generate outputs such as predictions, content, recommendations, or decisions that can influence physical or virtual environments", and adds that such systems vary in their levels of autonomy and adaptiveness after deployment (OECD.AI).

Where the name comes from. The term "artificial intelligence" appears in the title of a research proposal dated 31 August 1955, written by John McCarthy and three colleagues, for a summer study at Dartmouth College in the United States in 1956 (McCarthy et al., 1955, Stanford copy).

Types and branches of AI

AI is usually divided in two ways: by capability, and by the field or method it uses.

1. By capability

  • Narrow AI. Systems designed for a specific task or set of tasks, such as voice assistants and chatbots (IBM; Britannica). Everything we use today is of this kind, even the most powerful systems. The International AI Safety Report 2026 describes their abilities as "jagged": they excel at some difficult tasks yet sometimes fail at simpler ones, such as counting objects in an image (executive summary).
  • Artificial general intelligence (AGI). The idea of a system that can understand, learn and apply knowledge across a wide range of tasks at a level equal to or beyond a human's. IBM describes it as "currently theoretical" (IBM), and Britannica notes that no program yet matches "full human flexibility over wider domains" (Britannica).
  • Artificial superintelligence (ASI). A hypothetical intelligence far beyond humans in almost every field. It does not exist today, and experts disagree on whether and when it could appear. We explain this calmly in our guide to general and superintelligent AI.

2. By field or method

  • Machine learning. A method in which the computer learns from many examples instead of being given fixed rules. It is the basis of most AI today (IBM; glossary).
  • Deep learning. A branch of machine learning that uses neural networks with many layers. It is especially suited to language processing and computer vision (IBM; glossary).
  • Natural language processing (NLP). Understanding and producing text and speech, including machine translation and speech recognition. One Gulf example is a test NVIDIA published in September 2026, in which the share of words transcribed wrongly in two Saudi dialects (Najdi and Hijazi) fell from 55.05% to 29.96% (NVIDIA).
  • Computer vision. Enabling a computer to recognise what is in images and video (IBM).
  • Generative AI. Tools that write text and create images on request, such as ChatGPT. We explain how to use it safely in our guide to ChatGPT in Arabic.

Benefits of AI: examples from the Gulf

The International AI Safety Report 2026, which focuses on risks, says these systems are "already being usefully applied in healthcare, scientific research, education, and other sectors, albeit at highly uneven rates globally" (executive summary). Here are examples from the region:

  • Health. Abu Dhabi's M42 launched Ossi, a platform for clinicians, on 22 October 2026. It says Ossi is already in use at seven of its facilities in the UAE and Bahrain; one of its tasks is to transcribe a consultation and prepare draft notes for the clinician to review (M42).
  • The Arabic language. Abu Dhabi's Technology Innovation Institute announced Falcon-Emirati on 6 October 2026. It was developed using native Emirati content and Modern Standard Arabic material on Emirati culture and heritage (TII).
  • Education. The UAE Ministry of Education announced AI as a subject in the national curriculum in public schools, from kindergarten to Grade 12, starting in the 2025–2026 academic year (WAM; Gulf News).

Risks and harms of AI: what the sources say

The International AI Safety Report 2026 divides the risks of general-purpose AI into three categories: malicious use, malfunctions and systemic risks (executive summary). Here are the main ones in brief, with our pages that look at each fear in detail:

  • Mistakes and made-up information. The report says current systems sometimes fabricate information and give misleading advice. So check any important fact before you pass it on.
  • Scams and deepfakes. The report says AI is being misused to generate content for scams, fraud and blackmail. → How do I spot an AI scam, and where do I report it?
  • Jobs. The report says early evidence shows no effect on overall employment, but some signs of declining demand for early-career workers in some AI-exposed occupations, such as writing. → Will AI take my job?
  • Bias. Systems learn from data that can leave out languages and cultures, Arabic among them. → Is AI biased against people like me?
  • Privacy and surveillance. → Will AI be used to watch us?
  • Relying on it too much. The report says early evidence suggests reliance on AI tools can weaken critical thinking skills and encourage "automation bias", the tendency to trust AI outputs without enough scrutiny.
  • Loss of control. The report says "Current systems lack the capabilities to pose such risks", but that they are improving in relevant areas. → Will we lose control of AI?

Both things are true: the risks are real and taken seriously, and work to reduce them is under way in the Gulf and around the world.

Who oversees AI in the Gulf?

Gulf states have their own AI bodies and strategies. In Saudi Arabia, the Saudi Data and Artificial Intelligence Authority (SDAIA) was established by Royal Order No. (A/471) dated 29/12/1440 AH (30 August 2019), and is linked directly to the Prime Minister (SDAIA; Saudipedia). To see who sets the rules in each Gulf state, read AI governance in the Gulf.

If you are writing a report or essay on AI

  1. Check your school's or university's rules first. In Sharjah, for example, the Sharjah Private Education Authority launched a framework in October 2026 that asks schools to set out how AI may be used in assignments, projects and assessments, with its use disclosed where needed (SPEA; our news item).
  2. Use a simple structure: an introduction and definition, then the types and branches, then the benefits with examples, then the risks, then a conclusion with your own view.
  3. Cite the original source. These are reliable sources we used on this page, and you can go to them directly:
    • The OECD definition of an AI system (OECD.AI).
    • The International AI Safety Report 2026 (executive summary).
    • The 1955 Dartmouth proposal (Stanford).
    • The UAE Ministry of Education's announcement of the AI curriculum (WAM).
  4. If you use an AI tool, check every fact and source it gives you. These tools can invent references that do not exist, and the international report warns about fabricated information.

What we don't know yet

  • Whether and when general AI will appear. Experts disagree widely, and there is no agreed definition or test for it.
  • How large AI's long-term effect on jobs will be. The international report says economists disagree.
  • How widespread harms such as AI scams are. The report says systematic data on their prevalence and severity remains limited.

What this means for you

AI is not a mind or a person. It is a set of tools that learn from data, and you already use some of them every day. It has real benefits, which we can see in health, language and education in the Gulf, and real risks that deserve attention without panic. The best start is to understand it, try it on simple things, and check what it tells you.

We checked the sources on 31 October 2026, unless stated otherwise.