Learn AI /Responsible AI, Human Rights and Safety /Digital divide
Digital divide
- Starter
- Responsible AI, Human Rights and Safety
- Policy officers
- Programme & data teams
Also called: connectivity gap.
Back to the milestone you came from
In five seconds
What it is
The gap between people who can reliably use digital tools and services and those who cannot, because of access, cost, skills, language, design or safety barriers.
In plain words
Some people can use digital services easily, and some cannot. The difference comes from things like connection, cost, devices, skills, language, accessibility and safety.
An everyday example
A bus operator replaces printed timetables with an app. One passenger checks the next bus immediately. Another has an older phone with no space for the app. A third can download it, but the timetable is not compatible with a screen reader. Someone else understands the route but cannot afford mobile data every day. The timetable exists for everyone in theory, yet only some passengers can use it reliably enough to arrive on time. The divide is not just owning a phone. It includes cost, design, skills and whether digital access produces the intended benefit.
A survey is one short task, but the digital divide affects many parts of life over time, such as schooling, work, health information and civic participation. The analogy also focuses on using a service, while the real issue includes creating technology, shaping standards, and deciding whose needs are considered in the first place.
Remember thisOffering a digital channel does not create equal access. Check who can reach it, use it safely and turn access into a real benefit.
Understand it
Understand it
Why it matters
The digital divide changes who is seen, heard and served. If a programme, policy or AI system assumes everyone has stable internet, a private device, enough data, the right language and the skills to use digital tools safely, decisions will favour some groups over others. That can produce unfair service delivery, misleading data, weaker consultation and higher risk for people already facing disadvantage. For non-specialists, the key point is practical: before relying on digital or AI systems, ask who can realistically use them, who cannot, and what non-digital or supported alternatives are needed.
How a digital divide turns into unequal outcomes
Show how barriers in access, affordability, skills, language and design can lead to lower participation and less fair outcomes.
01
Different starting points
People begin with different devices, connections, incomes, skills, languages and accessibility needs.
Draw a column of simple person icons with small labels: strong connection, weak signal, shared phone, low literacy, screen reader user, rural area.
02
Barriers appear
A service may require constant internet, a newer device, paid data, one language or a design some users cannot access.
Draw a second column of barrier cards between people and the service: cost, coverage, device limits, skills, language, inaccessible design, safety concerns.
03
Participation splits
Some users complete the task easily. Others abandon it, need help, or never start.
Draw arrows from barriers to three outcomes: completed independently, completed with difficulty, not completed. Make arrow thickness vary to show unequal numbers.
04
Outcomes become unequal
The final results under-represent some groups, which can distort decisions and make services less fair.
Draw a final column with icons for information received, service access, representation in data and decision quality; fade or reduce some icons for excluded groups.
A left-to-right flow diagram with four stages. Stage 1 lists people with different starting conditions. Stage 2 shows common barriers. Stage 3 shows different levels of digital participation. Stage 4 shows practical outcomes such as who receives information, services or influence. Read from left to right to see how small barriers accumulate.
Hover or tap each barrier to highlight which users it affects and which outcome changes.
Describe this without the diagram
This diagram explains a chain of cause and effect. First, people start from different situations: some have strong internet, some weak signal, some share a phone, some need assistive technology, and some have limited digital skills or different languages. Second, barriers in the service appear, such as high data cost, poor coverage, old-device incompatibility, inaccessible design or safety concerns. Third, participation splits: some complete the task easily, some complete it with difficulty, and some cannot complete it. Fourth, outcomes become unequal because the final information and services reflect the people who could participate most easily.
With reduced motion: A static four-column chart replaces animation. Arrows connect starting conditions to barriers, barriers to participation levels, and participation levels to outcomes, with colour and labels showing where exclusion increases.
A digital divide is an unequal ability to use digital technology in real life. The gap can start with basic access: no device, weak electricity, slow mobile signal, expensive data, or no broadband. It can also continue after connection exists, if people cannot afford enough data, do not have the needed skills, face language barriers, or use services that are not accessible.
What goes in are people with different circumstances. These include income, age, disability, location, education, gender, language, legal status and safety risks. Then digital systems are introduced: online forms, mobile apps, AI tools, remote learning, digital payments or information services.
What happens is that some people can use these systems easily, while others are partly excluded or pushed into harder, slower or more risky paths. A service may assume constant internet, a modern smartphone, literacy in one language, or the confidence to share data online. If those assumptions are wrong, the gap widens.
What comes out is unequal participation and unequal outcomes. Some people get information, services, jobs or a voice in decision-making sooner and more safely. Others may be missed, misrepresented or forced to depend on intermediaries. In AI work, this matters because systems built for the connected population may perform poorly or unfairly for everyone else.
The digital divide is often confused with the AI divide. They overlap, but they are not the same. The digital divide is broader: it includes ordinary internet and device access, digital skills and safe use. The AI divide focuses more specifically on who can build, use or benefit from AI systems.
How “The timetable that only some passengers can read” maps onto the real thing
| In the story | In the AI system |
|---|---|
| Residents with different phones, data plans and needs | People facing different levels of digital access, affordability and accessibility |
| A survey that assumes a fast connection and one design style | Digital services built with hidden assumptions about users |
| Missing reports from some neighbourhoods | Unequal participation that produces incomplete or biased outcomes |
A worked example
At a regional hub, a programme officer is finalising a monitoring report before clearance. An implementing partner has moved registration for a training programme to a mobile portal, but facilitators report that interested participants are being missed.
Starting from: Aggregated portal records, hotline logs, facilitator notes and accessibility feedback, with personal details removed.
- The officer compares completed registrations with reports of attempted registration through the hotline and facilitators, rather than treating portal completions as the full level of demand.
- The review finds first-level barriers, including shared devices and unreliable connectivity, and second-level barriers, including a data-heavy form, limited language support and poor compatibility with assistive technology.
- The partner introduces a lightweight form, telephone and in-person registration, translated guidance and accessibility testing.
- The monitoring report distinguishes online uptake from overall participation and records which groups still face barriers to receiving the programme's benefits.
Result: Low portal completion was partly a measure of access and usability, not low interest. Using several registration channels produced a more credible picture of demand and reduced the risk that digital access would determine who could participate.
In a United Nations workplace
At a country office town hall, a reporting officer is presenting a progress report due for clearance by close of business. An implementing partner asks why an online consultation is described as representative when facilitators received substantial feedback on paper from people who could not use the portal.
The portal records who could reach, understand and complete it, not everyone who wished to participate. Device sharing, data costs, language, accessibility, digital skills and trust can all shape whose views enter the file. The reporting officer should account for offline submissions and disclose the limitations of the digital channel.
Before you try thisDo not treat an absence of online responses as an absence of views or need. Personal or protection-sensitive comments must not be copied into an AI tool that is not approved for such material. Offline and assisted channels also need safeguards so contributors are not identified or exposed.
Where you may hear it
- During clearance of a digital public-service concept note
- At a vendor demo in a procurement meeting
- In General Assembly informals on digital inclusion
- During an accessibility review before a portal goes live
Avoid confusion
Avoid confusion
Often confused with
AI divide
The unequal ability of people, organisations or countries to build, access, use or shape AI systems and their benefits.
Accessibility
Designing tools, content and services so people with different disabilities can use them effectively, safely and with dignity.
Inclusion
Inclusion means designing and using AI so different people can access it, be represented in it, and avoid being unfairly left out or harmed.
Common misconceptions
Common misconception: It only means some people do not have internet access.
Correction: Access matters, but the gap is wider than connection alone. People may be technically online and still be excluded by data cost, device limits, disability barriers, low digital skills, language mismatch, harassment risks or lack of trust.
Common misconception: Once most people have smartphones, the problem is basically solved.
Correction: A phone does not guarantee meaningful use. An older device, expensive data, poor battery charging, unsafe shared access, inaccessible app design or forms that require high literacy can still block participation.
Common misconception: This is a technology issue, not a human rights or fairness issue.
Correction: Digital access shapes access to information, services, participation and remedy. When important systems move online, unequal digital access can reinforce discrimination or exclusion, so the issue affects fairness, inclusion and accountability.
Quick check
Which example best shows a digital divide rather than only a technical glitch?
The digital divide is about unequal ability to access and use digital systems in real conditions. The second option combines affordability, accessibility and design barriers that exclude part of the population, which is the core issue.
Go deeper
Go deeper
The simple version
Some people can use digital services easily, and some cannot. The difference comes from things like connection, cost, devices, skills, language, accessibility and safety.
The technical version
The digital divide refers to uneven capacity to access, afford, use and benefit from digital technologies and digital public or private services. It includes first-level access gaps such as electricity, network coverage, devices and bandwidth; second-level gaps such as digital literacy, language support, accessibility and trust; and outcome gaps, meaning differences in who can convert digital access into real social, economic or civic benefits. In AI contexts, the divide matters at several stages: whose data is captured, who can interact with AI-enabled services, who receives the benefits, and whose needs are under-represented in design, evaluation and oversight. It is related to inclusion and accessibility, but it is not identical to either one. Accessibility focuses on whether systems work for people with disabilities. Inclusion is broader and may include social participation beyond technology. The digital divide also differs from the AI divide, which more specifically concerns unequal access to AI capabilities, infrastructure, expertise and benefits.
Reviewed external resources
Primary sourceRecommendation on the Ethics of Artificial Intelligence (opens in a new tab)UNESCO
The UNESCO Recommendation gives a formal policy and ethics context for inclusion and equitable access, and suits policy and programme readers.
Intermediatelink checked 2026-07-31
Primary sourceB-Tech Project (opens in a new tab)UN Human Rights Office (OHCHR)
The OHCHR B-Tech Project offers a human rights lens on technology impacts and accountability, and suits readers working on governance or safeguards.
Intermediatelink checked 2026-07-31
EncyclopaediaDigital divide (opens in a new tab)Wikipedia
Wikipedia gives a broad overview of the term, its history and related ideas, and suits readers who want a quick starting point.
Intermediatelink checked 2026-07-31
VideoVideo explanations: Digital divide (opens in a new tab)YouTube search
This YouTube search results page helps readers find a range of video explanations at different levels, and suits people who prefer to learn by watching; it is a search page, not one vetted source.
search results, not one pagelink checked 2026-07-31
DiscussionCommunity discussion: Digital divide (opens in a new tab)Reddit search
This Reddit search results page helps readers see how people discuss and question the topic in informal language, and suits readers exploring public conversation; it is a search page, not one authoritative source.
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
Offering a digital channel does not create equal access. Check who can reach it, use it safely and turn access into a real benefit.