Izabela Misiorny, Chief Marketing Officer of Siteimprove, dives into the impact of AI on web accessibility, revealing the challenges, solutions, and the need for human-centered models to foster inclusive digital spaces.
Everywhere you look online, you’re hit with dramatic headlines about how digital innovation will either be our savior or our downfall.
It’s true: digital technology has changed the way we socialize, work, shop and live. As the proliferation of AI (artificial intelligence) becomes more prominent, it will continue this trajectory at a rapid pace. But by describing technological developments as having a life of their own, we eliminate what is at their heart: people.
Ultimately, technology is created by people to help them.
And that’s the case. For people with disabilities, for example, technology has made it easier to participate in digital life and access important products, services and information.
However, there is still work to be done to ensure that the digital world is inclusive for all and that we all have equal access to the information we need and the experiences we deserve.
AI-based tools are being created and adopted at an exponential rate. For marketing teams tasked with creating digital experiences, there are a few things you need to know to use AI responsibly while ensuring everyone can access your digital experiences.
Putting people first in an AI-powered world
We already have a warning about what can happen when technology causes more harm than it helps.
Take accessibility overlaysFor example.
Several products on the market claim to make your website accessible using a toolbar, plugin, app, or widget. At first glance, this seems harmless and even quite ideal for businesses that want to offer an accessible website version quickly and without the necessary time, costs and accessibility expertise.
The problem is that these tools are superficial fixes that don’t fix accessibility issues in your website’s source code and, therefore, don’t provide a truly accessible experience or protect you from lawsuits.
The same thing is happening now with AI and accessibility. Several AI-based accessibility tools are quickly emerging. This is not a bad thing in itself. But let’s assume that the technology developers and marketers who buy them don’t do their due diligence to ensure that these products create value. In this case, we risk cluttering the market with products that create inaccessible websites.
AI does not solve all problems on its own
Let’s not forget that AI still has some issues related to biased datasets that can affect the results.
Furthermore, this does not absolve marketers of their responsibilities. Generative AI tools cannot be held accountable and, when used alone, do not take into account the wide range of perspectives needed to ensure your content is truly accessible.
The key to success will be integrating these AI tools into your existing workflows, with all the checks and balances you need to ensure your experiences are on-brand, compliant and fully inclusive.
So what does this mean for marketers?
We need to think about how to break down barriers between the tools in our arsenal and re-evaluate how software should work seamlessly (and in service of the overall user experience).
This is a challenge to face when working on the integration between a website and a CMS (Content Management System). Focus on bringing optimization capabilities together to save time, drive automated recommendations, and ensure consistency for the end user across all their digital experiences.
And the same goes for AI tools. The shortcomings of generative AI can be mitigated with solutions that prioritize human ethics, values, and collaboration. The challenge now is to ensure we create processes that leverage the speed and efficiency of AI without losing these essential attributes.
See more : How AI and ML impact content marketing and SEO
Balancing AI benefits with accessibility performance
Most accessibility experts agree that automated software can only flag about 30% of accessibility issues. The rest needs to be evaluated by humans. Software understands what is written better; there are simply too many contextual factors surrounding accessibility for our current level of technology. And it looks like that will be the case for a while.
What AI can do, however, is detect simple problems at scale. This does not replace the benefits of manual testing by experts. Equipping your marketing team with AI-powered tools can help you do more with the resources you have and create an effective baseline for web accessibility.
Here are some ways you can use AI to help you generate content faster while providing that essential human touch that keeps things socially conscious, unbiased, and inclusive.
Write image alt text
Whenever your content includes an image that provides additional context, you should include alt text (or alt text for short) in the image source code. This is so that people using screen readers can also understand what is displayed in the image. Plus, it’s important for search engines.
A handful of tools have already been developed specifically for this purpose. Using computer vision and recognition, these tools can evaluate an image and identify objects in the image. Some AI image recognition tools also offer an API key that your web team can integrate directly into your CMS.
This exciting initiative could potentially enable content teams to create more accessible content. As long as you don’t forget that these tools are not infallible and still require a human to examine the results.
Use inclusive language
AI-generated content should always be checked for biased or discriminatory language.
If you’re just starting to write accessible and inclusive content, you may not know what to look for. And it will take some time to acquire this knowledge, as our thoughts on inclusive language and social consciousness are constantly evolving. But some tools can help you develop a built-in check for inclusive language; you can ensure that any AI-generated content you distribute is scanned for bias and discrimination.
Automatic generation of video subtitles
Auto-generated captions have been around for a while. Despite this, they have a reputation for being inaccurate.
With the help of AI and machine learning, speech-to-text transcription is improving. But just like an image’s alt text, they still need a human to verify that the captions match the spoken words.
Still, auto-generated captions are a definite time saver, and their convenience encourages teams to use them consistently, not just on websites but across all media channels.
Writing descriptive video transcriptions
In addition to captions, descriptive video transcripts are also important for making videos accessible. A descriptive transcript includes a description of all visual content, including any on-screen text, graphics, people, settings, or other details.
Many teams struggle to commit to writing sufficiently detailed descriptive transcripts. With the rapid development of computer vision, this is an exciting possibility to watch for in the near future.
Innovate through accessible thinking
Building AI solutions with accessibility in mind has the potential to spark the next innovation. But it’s when you look at the possibilities of new technologies and combine them with the best practices we’ve developed over the past decades that you can start delivering better experiences today. Leveraging the limitless opportunities of AI, then applying the human perspective, brings things back to what matters: the users.
If you look at the features we commonly use, such as touchscreens, autocomplete, subtitles on videos, voice search, and speech-to-text on our smartphones, these are features that are specially designed for people with disabilities. And because they have improved the way people can use technology, they have been widely adopted by the rest of society.
Whatever the next innovation in accessible web experiences, you can stay ahead of the curve by ensuring your toolkit has the flexibility to adapt and the context to anchor it.
What AI-related tool have you used this year to make your workplace accessible and inclusive for your employees? Let us know on Facebook, XAnd LinkedIn. We would love to hear from you!
Image source: Shutterstock
