Balancing Accessibility, Security and AI: Design Inclusive Security Tools
summary
Claims that AI plus inclusive design makes security usable for disabled defenders, with concrete demos: a blind analyst runs code in a screen-reader-accessible Jupyter notebook, trains a simple anomaly-detection model on baseline logs and flags 693 anomalies which are filtered by new destinations and max outbound traffic to yield ~20 device pairs for investigation; a prototype multimodal auth uses voice + authenticator TOTP exposing timeout and usability tradeoffs; recommends WCAG, captions, audio description, early user inclusion and shifting accessibility left.
AI-written summary · about
chapters (12)
- 0:00Welcome
- 1:50Security incidents and AI tools
- 3:15Breaking cybersecurity stereotypes
- 4:26Accessible log analysis demo
- 13:51One-size-fits-all security problems
- 18:49Inclusive authentication with AI
- 23:08Accessibility as security superpower
- 26:21Designing accessible security tools
- 31:43Accessibility in development
- 33:51Disabled talent in cybersecurity
- 36:22Accessibility drives innovation
- 37:23Q&A
related talks
48mProgramming for accessibility
48mAI and Disability: Progress or Exclusion?
52mPreparing Accessibility for the Future and the Present
51mThe Intersection Of Accessibility And Inclusive Design
34mAI and Accessibility: the Good, the Bad, and the Bollocks
16mAccessibility: More than Just Screen Readers and ARIA
49mUsing AI to Streamline Inclusive Design Without Losing Humanity
45mAccessibility Beyond Code Compliance