Google researchers say they are shifting their accessibility work from reactive tools to agentic systems that are native to the interface, after identifying what they call the “accessibility gap” — the delay between the release of a new feature and the creation of an assistive layer for it.
In the research described, multimodal AI tools are presented as a path to more accessible interfaces. One prototype for web readability uses a central Orchestrator as a strategic reading manager, maintaining shared context and delegating work to expert sub-agents instead of forcing users through complex menus.
The system includes a The Summarization Agent: that breaks down complex documents and delegates key tasks, and a The Settings agent: that handles UI adjustments, such as scaling text, dynamically. The research says this modular approach lets users interact more intuitively and helps ensure specialized tasks are handled by the right expert without users needing to find the “correct” button.
The team also said it is working toward multimodal fluency beyond basic text-to-speech. Using Gemini’s ability to process voice, vision, and text simultaneously, the researchers have built prototypes that can turn live video into immediate, interactive audio descriptions.
According to the research, this is intended to improve situational awareness. In co-design sessions, the team observed that letting users query their environment in real time for specific visual details can reduce cognitive load and turn a passive experience into an active, conversational one.
Source: research.google.
Companies can share verified announcements through Newz9’s international press release submission page.
