Between November and June this year, our Assistive Technology Network (ATN) Leadership Group ran four ‘Community Conversation’ discussion sessions. These gave members a dedicated space to share experiences, challenges and ideas around assistive technology in further and higher education.
During these sessions, we split into groups based on the areas that interest us the most. We had an impressive range of interesting topics and over the year we’ve covered:
- Smart glasses as Assistive Technology
- Supporting visually impaired (VI) students to be ready for FE/HE
- The impact of generative AI on assistive tools
- Identifying, finding and accessing AT training and support
- Providing live captions
- AT and assessment considerations
- Researching Assistive Technology
- Free vs Paid assistive tools
- AT: The practicalities of balancing innovation with compliance?
Below we’ve rounded up the discussions from four of the most popular discussion topics from across the year: live captioning, free vs paid AT, generative AI, and AT training.
Interested in joining a future session? Invitations are sent to the ATN Jiscmail list which is open to all working at FE and HE institutions in the UK, join here with your institutional email address.
Experiences of Providing Live Captions
This discussion considered the opportunities and barriers to providing live captioning in education. While many members had extensive experience of captioning recorded content, live captioning was less widely adopted and presented a different set of considerations.
A key theme was the broad range of students who can benefit from live captions. Alongside deaf students, members highlighted the value for learners who experience auditory processing delay and those with English as an additional language.
Accuracy remains one of the biggest challenges. Members discussed issues with automated captioning, including latency and difficulties handling different accents, speech patterns and specialist terminology. A point was made that in some cases, bad captions are worse than having none at all.
Several members also highlighted challenges around approval and adoption of captioning tools. Some tools which provide live captioning also include note-taking and generative AI features, such as automatically producing summaries or key points. This raises familiar concerns around copyright, privacy and academic integrity, echoing the conversations taking place around generative AI in AT.
Though barriers are significant, there was a feeling that captioning tools themselves are improving. There were a range of experiences of both standalone and integrated tools in the space, including integrated tools such as PowerPoint Live Captions and macOS Live Captions, alongside dedicated services including Caption.Ed and Caption Connect.
The conversation also highlighted the importance of personalisation. Features such as displaying captions on an individual’s own device, allowing users to customise font, colour and text size, and speaker identification were all recognised as valuable. The customisation options available within macOS Live Captions were highlighted as an example of good practice. Some tools now provide speaker identification too, but it was felt that it is difficult to find tools that could do this consistently and accurately.
Free versus Paid Assistive Technology
In this discussion, we reflected on the differences between free and paid AT tools, where each can be valuable and where the negatives may outweigh the benefits. This topic was put forward in part due to the DfE Disabled Student’s Allowance (DSA) consultation, which suggested in its proposals that in some software categories free to access assistive tools can provide adequate support for disabled students.
Members agreed that the growing number of accessibility features built into mainstream software has been largely positive, making many forms of support more readily available without additional cost. At the same time, there was clear recognition that specialist assistive technology remains essential for many students.
The discussion also considered how institutions support students who are not eligible for DSA or other funding. Some members look to offer that support to those students directly, making a combination of free and paid tools available to students. They noted that this support was particularly well taken up by international and apprenticeship students. However, rising costs of specialist software and increasing pressure to demonstrate value for money were both highlighted as growing challenges.
Interoperability also emerged as an important consideration. Paid tools were generally seen to be more able to be used across different platforms and systems. Whereas free-to-access integrated tools are typically restricted to their platform. Members also reflected that, historically, smaller assistive technology companies have been more responsive to feedback than the larger mainstream technology providers.
The discussion naturally touched on the DSA process itself. Members recognised that students can sometimes be recommended a large number of tools which go unused, creating confusion rather than support. The importance of effective AT training was raised, to ensure that students are actually able to make use of the tools they are given. It was felt that there could be improvements made to the needs assessment process and guidance given, rather than limiting the specialist tools that are available through DSA which would disadvantage the students who need them.
Generative AI in AT
This was one of our most popular topics of the year, as such a considerable number of assistive tools now incorporate generative AI features.
Members shared many examples where these additions have significantly enhanced existing tools. Picture Smart within JAWS, which uses generative AI to describe images, and automatic lecture summaries and notes through tools such as Jamworks were both highlighted. At the same time, members have found it challenging to keep up with changes to existing tools. Particularly as they often integrate generative AI in different ways and are continually changing.
This creates particular challenges around assessment. Members discussed the uncertainty experienced by disabled students who rely on assistive technologies that now include generative AI functionality. Blanket restrictions on generative AI can unintentionally disadvantage these students where guidance does not clearly recognise reasonable adjustments or the role of assistive technology.
There was also discussion around the confusion that can exist between generative and non-generative AI. Members felt this can sometimes lead to tools being restricted unnecessarily because staff are unsure how they work or where generative AI is being used
Copyright was another important theme. Generative AI tools can be highly effective at reformatting or transforming materials to meet different accessibility needs. However, doing so often requires uploading content into third-party tools, making it important to understand how those services store and use uploaded material.
A clear recommendation from the discussion was that institutional AI related policies should explicitly address assistive technology. Doing so can provide reassurance for disabled students while also helping staff understand where AI-supported assistive technology fits within existing guidance.
These conversations have since led to the creation of a Generative AI in Assistive Technology Working Group, which is developing resources to support institutions in creating effective policies for generative AI-enabled assistive technology.
Finding and Accessing Quality AT Training
These discussions considered AT training for staff, students and learners. How people find training opportunities and what challenges there are to accessing training.
Some members were responsible for creating and delivering AT training in-house at their institutions. They noted challenges offering training for subscription-based tools, as licensing restrictions could prevent staff from accessing the software needed for training. Some reported low uptake on AT training but it was felt that this likely comes more from a lack of awareness of training rather than a lack of need.
Security and data protection requirements were another recurring theme. Some useful assistive tools cannot be approved for institutional use because of security concerns, creating a risk that staff instead turn to unapproved alternatives. Similarly, staff receiving software through Access to Work, have found that they could not use them within their institution due to data security concerns.
The discussion also considered how good digital accessibility practice can be supported across an institution. Student‑led accessibility initiatives were highlighted as particularly effective at influencing staff practice, with Accessibility Allies at the University of Southampton being referenced as a great example. Other recommendations included creating accessible templates in collaboration with branding/design teams, and providing open-source AT tools (e.g., NVDA) to staff software libraries so they can be used for testing.
These conversations also touched on the DSA process, and the importance of quality training for students on the tools they are recommended. It was noted that the transition period moving into HE is a point where quality training is particularly important. Some students may receive new AT shortly before, or even after, starting their course, making timely training essential if those tools are to make a meaningful difference from the outset.
Join a Community Conversation next year
We’ll be back next academic year with another series of regular AT Community Conversations.
Sessions are open to all working at FE and HE institutions in the UK, join here with your institutional email address to receive invitations.
You can also join Jisc’s Accessibility Community to discuss all things accessibility, including a dedicated Teams space and monthly drop-in sessions.