Learn AI /Data and Infrastructure /Cloud computing
Cloud computing
- Starter
- Data and Infrastructure
- Programme & data teams
Also called: the cloud.
Back to the milestone you came from
In five seconds
What it is
Using on-demand computing services over a network, instead of relying only on machines you manage directly.
In plain words
Instead of buying and running all your own computers, you rent computing services from a provider over a network and use them when needed.
An everyday example
A small team in Nairobi needs a short field report translated into French today. Instead of hiring full-time translators for every language, they send the text to a trusted translation service centre that already has staff, desks, dictionaries, and workflow tools. The centre does the work and sends the translated report back quickly. If ten more reports arrive tomorrow, the team can ask for more help without building a bigger office first.
A translation office uses people, while cloud computing mainly uses automated systems running on machines. Also, cloud services are not always as simple as handing work to one centre: data may be copied across regions, split across many servers, or controlled by contracts and technical settings. The analogy helps with the idea of renting capacity, but it does not explain networking, security controls, or data location well.
Remember thisCloud computing provides remote software and infrastructure on demand. The provider runs the underlying systems, but the office remains responsible for access, data handling, cost and continuity.
Understand it
Understand it
Why it matters
Cloud computing affects cost, speed, risk, and control. A team can launch services faster because it uses managed infrastructure instead of buying and maintaining every machine. That can help with scaling up during busy periods and scaling down when demand falls. But convenience can hide important choices. Where the data is stored may affect legal obligations and data sovereignty. How the service is configured affects security, access control, backup, and reliability. Pricing can look simple at first but grow with heavy use, storage, or data transfer. For non-specialists, the key decision is not only “cloud or not”. It is which service, for which data, under which rules, with what oversight, and at what total cost.
How cloud computing works in practice
After viewing this diagram, a reader should be able to explain how a user request goes to remote computing resources and returns as a result, and why this differs from running everything on one local machine.
01
1. User starts a task
A person uses a laptop or phone to store a file, open an app, or request a computation.
Draw a laptop and phone in the first box with short labels: file, app, task.
02
2. Request travels over a network
The task and any needed data are sent over the internet or another approved network connection.
Draw arrows from devices to a cloud icon or network line in the second box.
03
3. Remote systems handle the work
Servers in a provider’s data centre store data, run software, or process the request using shared resources.
Show stacked server racks in the third box with three small tags: storage, software, compute.
04
4. Result comes back
The completed file, answer, or service response is returned to the user’s device.
Draw an arrow back to a screen in the fourth box showing a completed document or status tick.
A left-to-right flow with four boxes connected by arrows. It shows a user device sending a task through a network to a cloud provider, where remote servers store or process it, then returning the result. Read from left to right.
Hover or tap each step to reveal a one-sentence note about what goes in, what happens, and what comes out.
Describe this without the diagram
The diagram is a four-step flow read from left to right. Step 1 shows a user on a laptop or phone starting a task such as saving a file, opening software, or requesting a translation. Step 2 shows that request moving over a network connection. Step 3 shows remote servers in a provider data centre handling the work, labelled storage, software, and compute. Step 4 shows the result returning to the user’s device as a finished file, response, or running service. The main message is that the computing work happens on remote infrastructure rather than only on the user’s own machine.
With reduced motion: A static four-box flow replaces any animation. Arrows remain fixed, and each box contains its label and explanation in visible text.
Cloud computing means accessing on-demand computing services over the internet or another network. Those services can include storage, databases, software, and processing power. Processing power means the capacity of processors to perform computations.
What goes in can be your files, your software, or a request to run a task. For example, you might upload a document, send data to be analysed, or ask an application to translate text. The request travels to remote servers, which are computers in large data centres run by a provider.
What happens next is that the provider’s systems store the data, run the software, and send back a result. You do not usually see the physical machines. You mainly choose what service you need, how much you need, and who is allowed to access it.
What comes out can be a saved file, a web app, an API response, or a completed computation. The main benefit is flexibility: you can start quickly and increase or reduce usage without buying new hardware. But cloud computing is not just “someone else’s computer”. That phrase is useful, but too simple. The details of security, location, cost, reliability, and legal control still matter.
How “Using a shared translation office” maps onto the real thing
| In the story | In the AI system |
|---|---|
| translation service centre | cloud provider running remote servers and services |
| sending the report out | sending data or a request over a network |
| centre’s staff and tools | shared computing resources such as storage, software, and processors |
| translated report returned | processed output returned to the user |
A worked example
A programme officer in Rome must prepare 25 cleared side-event concept notes as searchable files before the package goes for clearance at close of business.
Starting from: Twenty-five non-sensitive word-processing files uploaded to an approved cloud document service from the officer’s laptop.
- The officer signs in with an organisational account and requests conversion and indexing through the service’s web application.
- The provider authenticates the account and assigns storage and processing capacity from its remote, pooled infrastructure.
- The managed application converts the files, creates a search index and stores the outputs in the authorised workspace.
- The service returns links to the completed files, records the usage and releases processing capacity that is no longer needed.
Result: The clearance package is ready in a shared workspace without the programme officer installing conversion software or the office operating its own server.
In a United Nations workplace
At the annual workplan review, an internal auditor in a liaison office sees that three planned assignments will require temporary storage and processing capacity. Before the workplan goes to the front office for clearance at close of business, the auditor confirms that an approved cloud service can scale for the evidence files without purchasing a local server.
The provider operates the underlying servers, storage and software, then allocates capacity when authorised users request it. Pooled resources, remote administration and measured usage let the liaison office increase or reduce capacity as the audit workload changes.
Before you try thisAudit evidence may contain personal, HR, procurement-sensitive or security-sensitive material. It must not be uploaded to a cloud service that has not been approved for that material, because access, data location, retention and provider handling may expose protected information. Confirm permissions, logging, retention and secure deletion before use.
Where you may hear it
- A vendor demonstration during a procurement meeting
- An information and communications technology governance board reviewing a hosting proposal
- A country office business continuity exercise
- A budget review before a software subscription is renewed
Avoid confusion
Avoid confusion
Often confused with
Model hosting
The systems and services that keep an AI model available so it can receive requests, run on computing hardware, and return outputs reliably and securely.
Deployment
Deployment means putting an AI system into real use so people or software can send it live inputs and receive live outputs.
Edge AI
Running AI inference on a device at or near the data source, instead of sending each task to a remote server, which can reduce delay and network dependence.
Common misconceptions
Common misconception: The cloud means data floats around with no physical location.
Correction: Cloud services still run on physical machines in real data centres. The provider manages them, but location still matters for law, security, latency, and data governance.
Common misconception: Using the cloud is always cheaper.
Correction: It can reduce upfront hardware costs, especially for small or changing workloads. But ongoing charges for storage, computing time, users, and data transfer can become high if services are not planned and monitored.
Common misconception: Cloud computing and AI are the same thing.
Correction: They are different. Cloud computing is a way to provide computing resources over a network. AI may use cloud resources, but many cloud services have nothing to do with AI, and some AI runs outside the cloud.
Common misconception: If a provider is reputable, security is fully the provider’s job.
Correction: Providers handle some parts, such as operating the infrastructure, but organisations still need to manage their own accounts, permissions, settings, and safe use. Responsibility is shared, not transferred completely.
Quick check
Which description best matches cloud computing?
Cloud computing is about where computing resources are provided and managed: remote services over a network. It does not automatically mean AI, local-only storage, or a specific data location.
Go deeper
Go deeper
The simple version
Instead of buying and running all your own computers, you rent computing services from a provider over a network and use them when needed.
The technical version
Cloud computing is an on-demand delivery model for remote infrastructure and software services over a network. A provider operates data centres and exposes resources such as virtual machines, storage, databases, and managed applications. Users send requests from client devices; the provider authenticates access, allocates compute, stores or processes data, and returns outputs. Key properties often include pooled resources, elastic scaling, measured usage, and remote administration. The term is commonly confused with model hosting, but model hosting is one specific use of cloud or other infrastructure. It is also commonly confused with edge AI, which moves some processing closer to where data is produced rather than relying mainly on central remote servers.
Reviewed external resources
Primary sourceAI Risk Management Framework (opens in a new tab)NIST
This link gives a risk-management lens for using digital and AI systems, which suits managers and practitioners thinking about governance and controls around cloud-based services.
Intermediatelink checked 2026-07-31
Primary sourceOECD AI Principles (opens in a new tab)OECD
This link gives high-level principles on trustworthy AI and accountability, which suits policy and programme readers who need the broader governance context for cloud-enabled AI use.
Intermediatelink checked 2026-07-31
EncyclopaediaCloud computing (opens in a new tab)Wikipedia
This link gives a broad overview of the term, common service models, and history, which suits general readers who want orientation first.
Intermediatelink checked 2026-07-31
VideoVideo explanations: Cloud computing (opens in a new tab)YouTube search
This link is a YouTube search results page, so readers get a range of video explainers with different levels of depth; it suits people who learn best by watching and comparing sources.
search results, not one pagelink checked 2026-07-31
DiscussionCommunity discussion: Cloud computing (opens in a new tab)Reddit search
This link is a Reddit search results page, so readers get community questions, examples, and debates about practical use; it suits readers who want informal discussion but should verify claims carefully.
search results, not one pagelink checked 2026-07-31
Where this came from
Drafted with a language model through an offline pipeline, checked against this site's schema, and every external link above fetched to see whether it answers. Treat it as a starting point and check anything that matters against the primary sources. How this site is built and what that means.
The one thing to remember
Cloud computing provides remote software and infrastructure on demand. The provider runs the underlying systems, but the office remains responsible for access, data handling, cost and continuity.