Listen to Post

Disclaimer: This article is for general information and awareness purposes only and does not constitute medical, legal, technological or professional advice. Accessibility requirements vary between individuals, and a tool that benefits one person may not be suitable for another. AI-generated information can also be inaccurate, incomplete or misunderstood and should not be relied upon where safety, healthcare, legal rights or other important decisions are concerned.

AI Accessibility Tools That Can Support Independent Living and Work

From visual assistance and voice control to speech generation, transcription, reading support and workplace productivity, artificial intelligence is creating new ways for persons with disabilities to communicate, work, study and live more independently.

Features, pricing, availability, device compatibility and geographical restrictions may change. Users should check the provider’s current information before relying on any particular service.

Artificial intelligence is often discussed in terms of automation, business productivity and the future of employment, but some of its most important applications may be found in accessibility.

For persons with disabilities, technology can sometimes provide an alternative route around barriers that have traditionally restricted independence.

Someone who struggles to use a keyboard may be able to dictate instead.

Someone who cannot easily see their surroundings may be able to use artificial intelligence to obtain a description of what is in front of them.

Someone experiencing speech loss may be able to type words that are then spoken aloud.

A student who struggles to process large amounts of written information may be able to have that information summarised, explained or read aloud.

The following examples demonstrate how AI and intelligent accessibility technology can potentially support disabled people at home, in education and in the workplace.

Be My Eyes and Be My AI

Be My Eyes was originally developed to connect blind and low-vision users with sighted volunteers through live video.

The platform has since expanded to include Be My AI, which uses artificial intelligence to analyse images and describe visual information.

Users may be able to photograph an item, document, room, product or other object and ask questions about what the AI can see.

Be My Eyes states that Be My AI can assist with tasks such as reading labels, understanding documents, describing surroundings and navigating visual information. Human volunteer assistance remains available where additional help is required.

Be My Eyes also offers desktop versions for Windows and macOS that can describe screens, images and documents.

Who could benefit?

Potentially:

• Blind people;
• People with low vision;
• People who have difficulty interpreting visual information;
• Employees dealing with inaccessible visual interfaces;
• Students who require descriptions of visual material.

However, AI visual descriptions should not be treated as a replacement for a white cane, guide dog, mobility training or human assistance where physical safety is involved.

Microsoft Seeing AI

Microsoft’s Seeing AI uses artificial intelligence to describe visual information through a smartphone.

The application can help users identify and understand items such as:

• Printed documents;
• Handwriting;
• Products;
• Photographs;
• Colours;
• Currency;
• Surrounding visual scenes.

Microsoft describes Seeing AI as an application designed to narrate the world for blind and low-vision people and support everyday tasks such as reading mail, identifying products and understanding photographs.

Who could benefit?

Seeing AI could potentially assist people who:

• Are blind;
• Have partial sight;
• Find printed labels difficult to read;
• Need assistance distinguishing products;
• Need additional information about visual surroundings.

It may also provide greater privacy and independence where someone would otherwise need to ask another person to read a document or identify an item.

Google Lookout

Google Lookout is another application designed for people who are blind or have low vision.

It uses a device’s camera and sensors to identify text and objects and provide information about what the camera can see.

Depending on the feature being used, someone may be able to point their phone at information in their surroundings and have the device help interpret it.

Possible everyday uses include:

• Reading printed information;
• Recognising objects;
• Understanding signs;
• Examining documents;
• Identifying items around the home;
• Obtaining information about surroundings.

These tools can potentially reduce reliance on other people for straightforward visual tasks.

Voice Access for People With Mobility or Dexterity Impairments

Accessibility does not only relate to sight or hearing.

People living with paralysis, tremors, arthritis, neurological conditions, limb differences or restricted dexterity may find touchscreens, keyboards and computer mice difficult to operate.

Google’s Voice Access allows Android users to control their device using spoken commands.

Commands can be used to open applications, navigate screens, scroll, select controls and edit text hands-free.

Google specifically describes Voice Access as being useful for people who have difficulty manipulating a touchscreen, including because of paralysis or tremor.

Potential uses could include:

• Writing emails;
• Opening applications;
• Browsing websites;
• Composing messages;
• Navigating menus;
• Carrying out administrative tasks;
• Communicating remotely.

For some people, voice control could make the difference between needing physical assistance to operate technology and being able to use the device independently.

Apple Personal Voice

Losing the ability to speak can have a profound effect on independence and identity.

Apple’s Personal Voice enables eligible users to create a synthesised voice resembling their own voice.

Apple explains that Personal Voice can be used by people at risk of speech loss or those with speech disabilities. The created voice can then work alongside compatible communication features and applications.

The technology uses on-device machine learning and artificial intelligence to generate the personalised voice.

Who might benefit?

Potentially people living with conditions that affect or may progressively affect speech, including certain:

• Neurological conditions;
• Motor-neurone conditions;
• Muscular conditions;
• Speech disorders;
• Degenerative conditions.

The emotional significance should not be underestimated.

For someone facing progressive speech loss, being able to communicate through a voice that resembles their own may help preserve an important aspect of personal identity.

Apple Live Speech

Live Speech enables users to type what they want to say and have their Apple device speak the words aloud.

It can be used during supported phone and FaceTime calls, through compatible assistive communication applications and during face-to-face conversations.

It can also work alongside Personal Voice.

This could benefit people who:

• Cannot speak;
• Experience intermittent speech loss;
• Have speech disabilities;
• Find speaking extremely tiring;
• Use augmentative and alternative communication;
• Prefer typing during certain situations.

The ability to prepare commonly used phrases may also make everyday communication quicker.

AI-Powered Speech-to-Text and Live Transcription

For Deaf people and those who are hard of hearing, speech-to-text technology can make spoken conversations substantially more accessible.

Live transcription systems can turn spoken words into written text during:

• Meetings;
• Lectures;
• Webinars;
• Interviews;
• Appointments;
• Conferences;
• Online conversations.

Modern transcription services increasingly use machine learning to recognise speech, differentiate between speakers and generate searchable records.

However, transcription is not always accurate.

Accents, background noise, technical terminology and multiple people speaking simultaneously can all cause errors.

Automated captions should therefore be checked carefully, particularly in medical, legal, academic or professional settings.

Human captioning or British Sign Language interpretation may still be necessary depending on the individual’s communication requirements.

Generative AI for Dyslexia and Learning Differences

Generative AI assistants can potentially help people who have dyslexia, learning difficulties or other conditions affecting reading and written communication.

A user might ask an AI assistant to:

• Explain a complicated paragraph in plain English;
• Break instructions into smaller steps;
• Summarise long documents;
• Identify key points;
• Proofread spelling and grammar;
• Reorganise rough notes;
• Create revision questions;
• Explain unfamiliar terminology;
• Convert notes into structured headings;
• Draft information in a clearer format.

This can be especially useful where someone understands a subject well but struggles to organise their thoughts into written form.

However, students should always check the AI policies of their school, college or university before using generative AI in assessed work.

AI should support learning rather than replace it.

AI for Memory, Organisation and Cognitive Support

Some disabled people and people living with neurological conditions experience difficulties with memory, concentration, executive functioning or organisation.

Digital assistants and AI-powered productivity tools may help with:

• Creating reminders;
• Organising appointments;
• Breaking tasks into manageable steps;
• Creating checklists;
• Summarising meetings;
• Organising notes;
• Planning workloads;
• Prioritising tasks.

For example, instead of facing a vague instruction such as:

“Prepare everything needed for tomorrow’s appointment,”

Someone could ask an AI assistant to break the task into:

1. Find the appointment letter.
2. Check the appointment time.
3. Prepare identification.
4. Write down questions.
5. Gather medication information if needed.
6. Arrange transport.
7. Set a reminder.

Breaking information into smaller actions can sometimes reduce cognitive overload.

AI Reading Assistance

Reading large volumes of text can be exhausting or inaccessible for some people.

Text-to-speech technology can allow written information to be heard rather than read visually.

Modern accessibility systems can also:

• Read web pages aloud;
• Read emails;
• Pronounce selected words;
• Enlarge text;
• Simplify page layouts;
• Improve contrast;
• Reduce distracting content.

Microsoft, for example, provides built-in accessibility tools including Narrator, screen-reader support, Immersive Reader and text-to-speech functionality within Microsoft products.

Google Android also includes features such as TalkBack, Reading Mode and Select to Speak.

These tools may benefit people with visual impairments, dyslexia, neurological conditions or reading difficulties.

AI and Remote Working

AI accessibility tools could also expand opportunities for disabled entrepreneurs, freelancers and employees working remotely.

Someone who finds traditional workplaces inaccessible may be able to work from home using technologies that assist with:

• Dictation;
• Transcription;
• Research;
• Writing;
• Email management;
• Document summarisation;
• Meeting notes;
• Scheduling;
• Visual assistance;
• Proofreading;
• Communication.

Possible remote occupations could include:

• Content writing;
• Administration;
• Research;
• Customer support;
• Digital marketing;
• Legal research;
• Virtual assistance;
• Tutoring;
• Consultancy;
• Editing;
• Web development;
• Data analysis.

Employers should remember, however, that providing access to AI does not remove their responsibilities towards disabled workers.

Technology should complement reasonable adjustments rather than become an excuse for refusing them.

Alternative Controls and Switch Technology

Not everyone can use a touchscreen, keyboard or mouse.

Modern devices increasingly support alternative methods of interaction.

Android accessibility features include Switch Access, camera switches and alternative input methods alongside Voice Access.

These technologies can enable users to operate devices using movements, switches, keyboards or other alternative controls.

Combined with AI, such systems may become increasingly sophisticated, adapting interfaces around the physical abilities of individual users.

AI Image Description in the Workplace

Many professional environments rely heavily on visual information.

Graphs, dashboards, photographs, diagrams, website layouts and scanned documents may all create difficulties for someone with a visual impairment.

AI visual assistants can increasingly analyse information displayed on a screen and provide spoken or written explanations.

For example, Be My Eyes now offers desktop visual assistance capable of describing screen content, images and documents.

Microsoft’s Copilot Vision can also analyse a screen, application, webpage or supported camera view that a user chooses to share and answer questions about it, although availability and subscription requirements can vary.

Such technology could become increasingly valuable for workplace accessibility, although confidential information should never be shared with third-party AI systems without first understanding the organisation’s privacy and security requirements.

Smart-Home Technology

AI-enabled smart-home technology can also support independent living.

Depending on the equipment installed, users may be able to operate:

• Lighting;
• Heating;
• Televisions;
• Curtains or blinds;
• Doorbells;
• Security cameras;
• Appliances;
• Reminders;
• Music;
• Communication devices through voice commands, accessible smartphone controls or automation.

For someone with limited mobility, reducing the number of physical actions required around the home may make everyday life easier.

However, smart-home devices should never replace appropriate care arrangements, emergency support or properly adapted accommodation where those are required.

AI Cannot Replace Human Support

Technology can provide independence, but it should not create isolation.

There is a danger that organisations may see AI as a cheaper replacement for human assistance.

A disabled person should not be told:

“Use the chatbot instead. when what they require is accessible human customer service.

Nor should AI replace:

• Personal assistants;
• Carers;
• Interpreters;
• Healthcare professionals;
• Support workers;
• Teachers;
• Specialist disability advisers;
• Mobility training;
• Legal advice.

The strongest accessibility systems combine technology with the ability to obtain human assistance when required.

Be My Eyes is a good example of this model because AI visual assistance exists alongside access to human volunteers.

Privacy Must Be Considered

Some accessibility tools may process extremely sensitive information.

A user might ask AI to:

• Read a medical letter;
• Examine a bank statement;
• Summarise legal correspondence;
• Analyse an employment document;
• Describe their home;
• Listen to a meeting;
• Access a camera view.

Before uploading sensitive material, users should consider:

• Where the information is processed;
• Whether it is stored;
• Whether humans can review it;
• Whether it may be used for AI training;
• How it can be deleted;
• What permissions the application has.

Convenience should never require someone to surrender unnecessary amounts of private information.

Always Verify Important AI Information

Artificial intelligence can make mistakes.

An AI system might:

• Misread a medication label;
• Misunderstand a document;
• Describe an object incorrectly;
• Produce inaccurate captions;
• Confuse two products;
• Misinterpret a visual hazard.

Therefore, AI should never be the sole source of information when an incorrect answer could create serious consequences.

Medical information, medication instructions, legal documents, financial information and potentially hazardous situations should always be independently verified.

Accessibility Tools Should Be Affordable

Another challenge is cost.

Some accessibility features are provided free of charge or are built into smartphones and computers, while others require subscriptions, newer devices or additional equipment.

Accessibility should not become another area where independence depends entirely on someone’s disposable income.

Governments, educational institutions, employers and technology companies should consider how assistive technologies can be made available to people who cannot afford them independently.

The benefits of AI accessibility will be limited if the people who could benefit most are priced out.

Nothing About Us Without Us

The most successful accessibility technologies are likely to be those created with meaningful input from disabled people.

Developers should involve disabled users throughout:

• Research;
• Product design;
• Testing;
• Accessibility audits;
• Development;
• Feedback;
• Product updates.

Accessibility should not be based solely on assumptions about what developers believe disabled people require.

Persons with disabilities should have a direct role in shaping the technologies intended to support them. https://nothing-without-us.org/our-name-our-commitment/

Conclusion

Artificial intelligence cannot remove every barrier faced by disabled people, but it can help remove some of them.

AI-powered visual assistance can provide information that was previously inaccessible.

Voice control can help someone operate technology without using their hands.

Speech-generation tools can give someone another way to communicate.

Transcription can make spoken conversations accessible.

Generative AI can help organise information, simplify complicated text and support people in employment and education.

What matters is how the technology is designed and used.

The goal should never be to make people dependent on AI.

The goal should be to use technology to provide greater choice, greater independence and greater control.

For persons with disabilities, the most exciting future for artificial intelligence may therefore not be machines replacing humans.

It may be technology finally learning how to adapt to the diversity of the humans using it.

Further Reading & Resources

• https://unric.org/en/building-an-accessible-future-for-all-ai-and-the-inclusion-of-persons-with-disabilities/
• https://unric.org/en/building-an-accessible-future-for-all-ai-and-the-inclusion-of-persons-with-disabilities/https://nothing-without-us.org/our-name-our-commitment/
• https://www.seeingai.com/
• https://www.bemyeyes.com/bme-ai/
• https://support.google.com/accessibility/android/answer/9031274
• https://support.google.com/accessibility/android/answer/9158064
• https://www.apple.com/uk/accessibility/
• https://goblin.tools/
• https://speechify.com/
• https://otter.ai/
• https://chatgpt.com/