At the intersection of GenAI and application lifecycle — accessibility decoded

As companies strive to create inclusive digital experiences, Generative AI (GenAI) is anticipated to emerge as a transformative solution to bridge the accessibility gap. This innovative technology presents a compelling opportunity to impact accessibility initiatives in the future. By seamlessly integrating GenAI across the application lifecycle, companies can future-proof their offerings. This will not only guarantee compliance with evolving accessibility standards but also enhance the user experience for all.

The digital accessibility imperative

In an increasingly digital world, making sure access to information and services for individuals with disabilities is a strategic business imperative. After all, every sixth person on this planet experiences some form of disability. A failure to prioritize digital accessibility can result in legal and reputational risk, customer attrition and lost revenue opportunities. For example, a multinational pizza restaurant chain faced a lawsuit for not making its website and mobile app accessible to users with visual impairments. In addition, with an alarming statistic, WebAIM concluded that the average website had 56.8 detectable accessibility errors per page in 2024, underscoring the pressing need for companies to proactively address this issue.

Exploring the GenAI advantage: Unlocking value across the application lifecycle

Soon, GenAI will present a multifaceted value proposition for companies looking to enhance digital accessibility while perfecting resource allocation and streamlining processes. By leveraging GenAI, companies will unlock value across the entire application lifecycle:

1. Inclusive design and development: GenAI models will generate accessible UI/UX designs, wireframes and prototypes adhering to accessibility standards, such as the Web Content Accessibility Guidelines (WCAG). For instance, a GenAI model could generate multiple design variations that meet specific color contrast ratios and font size requirements for individuals with visual impairments. Additionally, trained models produce reliable and compliant code components, making sure accessibility is baked into applications from the ground up.
2. Testing and quality assurance: GenAI generates comprehensive accessibility test cases and scripts. Based on user scenarios and personas, these scripts could mitigate the risk of releasing inaccessible applications and provide thorough vetting for inclusivity. For example, a GenAI model could generate test cases that simulate how a screen reader would interact with an application, finding potential issues for users with visual impairments.
3. Continuous accessibility auditing and modernization: Regular accessibility audits are critical for maintaining compliance and consistent user experiences. GenAI models will be able to analyze code, websites and applications, finding potential accessibility issues and violations while suggesting remediation strategies.
4. Accessible content creation: Progressively, we are witnessing GenAI creating more accessible content, be it text, code or multi-media. However, GenAI models can generate extensive, accurate and descriptive alternative text (alt text) for images using the vision capabilities of large language models. This will enormously enhance accessibility and comprehension of visual content on the internet. Currently, one-third of the images on popular home pages have missing, questionable or repetitive alternative text. This capability can save valuable time and effort for clients.
5. Fostering an accessibility-driven culture: By generating tailored training materials, documentation and resources, GenAI empowers companies to cultivate a culture of accessibility awareness and best practices across teams and stakeholders. For example, a GenAI model could generate interactive training modules on web accessibility for developers, complete with practical coding examples and quizzes to reinforce learning.

Responsible implementation: Addressing ethical considerations

While GenAI presents significant opportunities, it is crucial to address the ethical considerations surrounding this technology. Potential biases in training data and outputs pose risks that must be mitigated through responsible implementation strategies. Key focus areas include:

• Prioritizing diverse and inclusive data sets: Training GenAI models on data sets that are representative of various demographics and free from societal biases is essential to safeguard unbiased and fair outputs. For instance, if a GenAI model is trained on alt-text descriptions written primarily from a western cultural perspective, it may struggle to generate correct and relevant descriptions for images from other cultural contexts.
• Maintaining human oversight: Human subject matter experts must review and confirm the outputs generated by GenAI to ensure accuracy and alignment with accessibility standards and best practices. This step is crucial given GenAI models may sometimes generate outputs that inadvertently perpetuate biases or have errors.
• Continuous monitoring and updating: Regular monitoring and updating of GenAI models is necessary to maintain their effectiveness and responsiveness to evolving technology and societal norms. As new accessibility guidelines are introduced, or user preferences change, companies should make sure their GenAI models still are up-to-date and aligned with these shifts.

Collaborative ecosystem for digital inclusion

Achieving true digital inclusion requires a collaborative ecosystem involving technologists, people with disabilities, accessibility advocates and industry experts. By fostering co-design, co-testing and active community engagement, companies can make sure that their GenAI solutions truly resonate with the needs of the intended users.

The path forward: A future-proof strategy

As the digital landscape continues to evolve and if done correctly, embracing GenAI for accessibility presents a future-proof strategy for organizations. By proactively implementing GenAI solutions and setting up responsible governance frameworks, companies can position themselves as industry leaders in digital inclusion, mitigating risks, enhancing customer experiences and unlocking new revenue streams. People with disabilities collectively stand for a consumer market of an estimated annual disposable income of over $490 billion in the United States alone.

Ultimately, the integration of GenAI into digital accessibility initiatives is not merely a technological endeavor but a strategic imperative. It aligns with broader organizational goals of innovation and corporate sustainability. By harnessing the transformative potential of GenAI responsibly and collaboratively, companies can pave the way for a more inclusive digital future.

At NTT DATA, we actively create inclusive spaces for people with disabilities. Our cross-functional effort aims to increase representation and support for individuals with disabilities.

Discover how NTT DATA’s Digital Application Services can help you with all your accessibility needs. We offer accessibility testing, remediation, Voluntary Accessibility Product Template (VPAT) certifications and much more.