Cephable

People with motor disabilities, ALS, cerebral palsy, and limited hand dexterity run Cephable on standard laptops to control computers and games through head movements, facial expressions, and voice commands when accessible input methods are absent from the software they need to use.

ENABLE Model location

• Builder-side Interventions → Develop Accessible Implementations
• Navigator-side Compensations → Use Assistive Technologies

What it is

Alex Dunn founded Cephable in 2021 as an AI-driven software platform that enables users to control any digital device through voice commands, facial expressions, head movements, body gestures, and virtual buttons configured through a companion mobile app. All AI inference runs fully on-device using the device's Neural Processing Unit, requiring no internet connection and no specialized hardware beyond the webcam and microphone built into most laptops.

Cephable operates at two distinct levels. As a standalone consumer application, it works as a navigator-side assistive technology that users install to compensate when applications and games lack built-in accessible input support. As a Software Development Kit (SDK), it enables developers to embed alternative input controls natively during the development phase, before a product ships. The SDK supports JavaScript, C#, C++, Unity, Unreal Engine, Python, iOS, and Android.

The consumer application comes pre-trained on hundreds of applications from Google, Microsoft, Apple, and other major software makers without requiring user configuration. Users with ALS, cerebral palsy, spinal cord injuries, and other conditions affecting fine motor control have used Cephable to pass bar exams in proctored environments, extend professional careers, control games competitively, and perform daily computing tasks that keyboard and mouse input had made inaccessible.¹ Brooke Eby, a Salesforce employee diagnosed with ALS in 2022, described the platform: "My ability to work has decreased a ton. Once my hand started getting weaker, it got harder for me to type and use a mouse. I'm so grateful that my accessibility team reached out about this technology because it allows me to work using voice commands, which will allow me to work for much longer into my diagnosis."²

Cephable employs a team in which 75% of members identify as having a disability and maintains a Consortium of approximately 2,000 community members with varying abilities who participate in continuous feedback from research and development through feature implementation, with power users compensated for their time.³ The company prices personal, non-commercial use at no cost, describing this as a deliberate policy to eliminate the "disability tax" that historically priced adaptive controls out of reach for many disabled users.⁴

Why it matters

Personal computing settled on keyboard and mouse as its default input interface before disability was a design criterion. The engineers and companies who built the first commercial personal computers in the late 1970s and early 1980s designed for an assumed able-bodied user. That assumption became structural. When Reg Maling prototyped the POSSUM (Patient Operated Selector Mechanism) at Stoke Mandeville Hospital in 1960, allowing tetraplegic patients to operate a typewriter through sip-and-puff controls, the innovation was treated as medical equipment rather than evidence that interfaces themselves should be designed differently.⁵ The ADA, signed into law in 1990, preceded the modern internet and addressed no digital content obligations directly, and legal interpretations of ADA applicability to websites and software remained contested for the next three decades.⁶ Section 508 of the Rehabilitation Act, amended in 1998, imposed requirement-setting obligations on federal agencies alone, leaving the commercial software and gaming industries without mandatory digital access obligations.⁷ By the time the International Game Developers Association established its first Game Accessibility Special Interest Group in 2003, the industry had shipped over two decades of games designed exclusively for two-handed keyboard and mouse or gamepad input.⁸ Mark Barlet founded AbleGamers in 2004 after watching his friend Stephanie Walker, who had multiple sclerosis, lose the ability to control a computer mouse.⁹ Microsoft launched the Xbox Adaptive Controller in 2018, the first major mainstream hardware AT device for gaming, as a late and partial acknowledgment of an exclusion built into the console architecture from the beginning. Cephable entered this sequence in 2023 as a software-only platform consolidating what users and organizations like AbleGamers called the "patchwork of assistive tech" that disabled gamers had been forced to assemble on their own.³

Game developers, software studios, and platform companies sustained the exclusion through professional norms that treated accessibility as a post-launch add-on rather than a design requirement. When the IGDA SIG conducted its first developer survey through Gamasutra in 2003, it found that accessible features appeared in games only sporadically and without design intention: closed captions in a title here, adjustable difficulty in a title there, but no systematic commitment to accessible input.⁸ Approximately 20% of casual video game players have a physical, mental, or developmental disability.¹⁰ Disabled gamers faced two simultaneous institutional failures: game studios that treated accessible input as optional, and an AT market that sold each piece of the solution separately at specialist prices. The result was a population for whom enduring inaccessibility or assembling expensive multi-device rigs was the only path to participation. Cairns, Power, Barlet, and Haynes (2021) documented what disabled gamers said they valued about accessible games: not restoration to a prior state, but the ability to play the same games as everyone else.¹¹ A survey of 230 AbleGamers volunteers found that disabled gamers played for health reasons, including stress management, depression, and physical therapy, at rates measurably higher than non-disabled players, and that social participation through gaming carried specific therapeutic weight that inaccessible design removed.¹²

The AT sector's pricing structure imposed an economic barrier that compounded the design barrier. A Tobii Dynavox eye-gaze communication device sells new for approximately $20,000; Dragon Professional, a voice recognition tool, costs $699 as a one-time license or $660 per year by subscription.¹³ A 2001 University of Michigan survey found that approximately 40% of AT users paid entirely out of pocket, with private health insurance and Medicare each covering only 15-20% of cases.¹⁴ The National Disability Institute's 2020 analysis found that households with a disabled adult require approximately 28% more income, $17,690 per year above median, to achieve the same standard of living as non-disabled households, and that accounting for these extra costs would reclassify 2.2 million more disabled people as officially poor.¹⁵ The WHO's 2022 Global Report on Assistive Technology found that globally only 1 in 10 people can access the AT they need.¹⁶ Marta Russell, in Beyond Ramps (1998) and Capitalism and Disability (2019), argued that AT pricing followed market design logic: the disability industry extracts surplus from disabled people's survival needs, turning impaired bodies into commodities around which profitable industries organize.¹⁷ Cephable's business model challenged this pricing arrangement directly. The platform offers personal, non-commercial use at no cost and shifts revenue to enterprise licensing, so the structural cost of accessibility falls on game studios and corporations rather than on individual disabled users who had been left to compensate for what developers chose not to build in. That inversion does not resolve the political economy of AT broadly, but it names the arrangement and builds against it within the commercial software sector.

Alison Kafer, in Feminist, Queer, Crip (2013), named the "curative imaginary" as the premise organizing most assistive technology design: "an understanding of disability that not only expects and assumes intervention but also cannot imagine or comprehend anything other than intervention."¹⁸ Alex Dunn began developing Cephable's predecessor technology after watching his younger brother, a person with a disability affecting fine motor control, struggle to keep up with friends in Minecraft during the early COVID-19 pandemic. The pandemic accelerated both the demand and the recognition: as digital education and remote work became mandatory, the gap between what keyboards and mice allowed and what disabled people needed became visible at institutional scale. Cephable's design reflects a partially different premise. The platform does not claim to fix disabled bodies but to make existing software controllable by the bodyminds that disabled users actually have. The Consortium model, where 2,000 community members with disabilities participate in R&D from design through implementation with power users compensated, builds the affected community into the production process rather than receiving it as a target population.³ Whether that represents a departure from the curative imaginary or a more sophisticated version of it depends on whether the platform's goal is normalization or genuine user-defined access. Ellis, Leaver, and Kent's Gaming Disability (2023) documented how gaming platforms historically produced "integrative access," tolerating disability without valuing it, and the scholarship suggests that individual company commitment, however sincere, does not substitute for structural requirements that make accessible development mandatory across the industry.¹⁹

As ALS advances, the keyboard-and-mouse default progressively excludes patients from professional communication, symptom self-reporting, and care coordination before it closes off game access. Londral (2022) found that 80-95% of ALS patients become unable to meet communication needs using natural speech at some disease stage, and that patients who retain digital AT access can continue self-reporting symptoms and participating in their own care decisions, while those who lose access become "disempowered to self-report symptoms and needs and increase their dependency on caregivers to make decisions."²⁰ Eby's documented experience at Salesforce demonstrates the mechanism: she uses Cephable to respond to messages, open applications, and complete work tasks as hand strength declines, and the social arrangement that defaults software interfaces to keyboard and mouse input produces a progressive loss of professional participation and communication capacity as motor disability advances. The bar exam case documented by EasterSeals Technology is structurally the same mechanism: Hogan Vansickle used Cephable to pass the North Carolina Bar Exam because its fully offline, on-device processing passed security systems that blocked internet-dependent AT, a barrier that would have excluded him from professional certification entirely.⁴ A 2024 PMC study of people with spinal cord injuries found that gaming "a valuable technical aid in enabling social participation in situations where physical gatherings are challenging," and that within virtual spaces "their physical disability is not visible."²¹ When developers choose not to build accessible input and users cannot afford the AT hardware, social participation and professional participation narrow simultaneously. Structural abandonment by software builders does not register as a policy choice; it registers in bodies as reduced work, reduced connection, and accelerating isolation.

Cephable consolidated voice, camera, and gesture input into a single application with a developer SDK for the first time, eliminating the multi-system assembly that previously required disabled users to buy and configure separate hardware and software for each input mode. Developers can embed alternative controls during the development phase rather than leaving users to compensate afterward. As of late 2024, Cephable reported late-stage conversations with major gaming publishers about bundling the SDK with new titles, with no signed agreements yet.²² Without regulatory requirements equivalent to Section 508 for commercial software, SDK adoption depends entirely on developer choice. The past two decades of voluntary accessibility commitments by game studios produced the patchwork that made Cephable necessary. Whether Cephable changes that pattern or becomes one more layer in it depends on whether the commercial software industry treats accessibility as a design obligation or leaves it, again, to individual users to solve on their own.

Real-world examples

In the news

How AI Is Improving Accessibility for Professionals With Disabilities (May 2025)
-- Jeffrey Howard, InclusionHub

• Brooke Eby, a Salesforce employee with ALS, uses Cephable's voice command platform to open applications, respond to messages, and complete work tasks as her hand strength declines. The piece documents how navigator-side assistive technologies enable workforce participation when motor disability narrows the range of standard input devices a worker can use. It also cites a finding from Salesforce research that 76% of individuals with disabilities choose not to disclose their condition at work, illustrating the social and professional costs of enduring inaccessibility in workplaces that have not built accessible tooling into their baseline infrastructure.

In the news

ATU706 – Cephable with Alex Dunn (December 2024)
-- Josh Anderson, Easterseals Technology

• The podcast documents Alex Dunn describing Cephable's origin: watching his younger brother, a person with a disability affecting fine motor control, struggle to keep pace with friends in Minecraft during the COVID-19 pandemic. Dunn cites cost as "the biggest barrier to folks getting the assistive technology that they need" and describes the decision to offer personal use at no cost as a direct response. The episode also documents the bar exam case: a user named Hogan Vansickle passed the North Carolina Bar Exam using Cephable because its fully offline processing passed security systems that blocked internet-dependent AT, then went on to practice disability rights law.

In the news

The latest video game controller isn't plastic. It's your face. (August 2022)
-- Amanda Florian, The Washington Post

• Florian documented the predecessor hardware device Enabled Play (renamed Cephable when Dunn pivoted to software-only), describing Alex Dunn, then 26, as a self-taught developer who built the device after watching his brother's gaming exclusion. Speech language pathologist Julia Franklin was quoted in the piece: "This allows individuals to leverage their strengths and movements that are already present." The article captures the moment when the underlying product premise shifted from hardware to software, foreshadowing the SDK-centered model Cephable later developed.

In the news

REPORT: Developer Spotlight – Cephable (May 2024)
-- Ben Bajarin and Max Weinbach, Creative Strategies

• Technical report on Cephable's integration with Qualcomm's Snapdragon X Elite Hexagon NPU, documenting that the platform achieved 30% less CPU and 50% less GPU usage compared to an M3 MacBook Pro while running camera-based alternative input controls. The report illustrates how builder-side development of accessible input is increasingly feasible on commodity hardware without performance trade-offs, undercutting the asserted feasibility barrier that studios and developers frequently invoke as a reason not to integrate accessible input earlier in the design process.

• Cephable won the SXSW 2024 Pitch Competition in the Innovative World category, selected from 800 applicants and 45 finalists.²³
• Rex Garrison, a real estate appraiser with ALS, uses Cephable voice commands for mouse movements across multiple monitors, extending both his career and financial independence as motor function declines.¹
• Hogan Vansickle used Cephable's fully offline voice and head-mouse controls to pass the North Carolina Bar Exam through security-monitored proctoring systems that blocked internet-dependent AT, then built a career as a disability rights attorney. He described the platform as "integral to my success."⁴
• Developer testimonials from Cephable's SDK trial page document integration times of three days to one week, with one gaming developer noting: "Using Cephable SDKs I had alternative controls integrated into our game in 3 days. I've never had something so easy to work with."²⁴
• Dayton Jones, a competitive e-sports gamer and streamer who uses Cephable, described the platform as a tool that could reduce stigma around AT: "Anybody can use it...to lower the stigma on assistive technology."¹
• By end of 2024, Cephable's user base had grown more than 10,000x from early adoption levels, with the company simultaneously in late-stage conversations with major gaming publishers about embedding the SDK in newly developed titles.²⁵

What care sounds like (builder-side interventions)

Care at the development stage involves building alternative input into a product before it ships:

• "We integrated the Cephable SDK during pre-production so that head movement and voice controls work at launch, not as a patch after complaints come in."
• "We scoped accessible input as a first-class feature in the game design document alongside standard controller mapping."
• "Our development sprint included testing the SDK on five-year-old hardware to confirm the performance overhead was acceptable. It was."
• "Before we shipped, we confirmed every interaction in the game (menus, combat, inventory) was fully operable by voice and camera controls, not just the first two tutorial levels."
• "We joined the Cephable Consortium to get feedback from disabled users during design, not after launch."

What neglect sounds like (builder-side interventions)

Neglect involves treating accessible input as optional, speculative, or not the developer's responsibility:

• "We'll add accessibility options in a post-launch update if there's enough player demand."
• "Our game is mouse-and-keyboard only. Users who need alternatives can find a workaround."
• "Integrating a third-party SDK isn't in scope for this release cycle."
• "We don't have data showing that players with motor disabilities are part of our audience."
• "That's an accessibility request. We don't have an accessibility team."
• "Players can already use the OS-level accessibility features. We don't need to build anything on top of that."

What compensation sounds like (navigator-side compensations)

Compensation describes the labor users undertake when developers have not built accessible input into the software:

• "I spend an hour every time there's a major update re-training my voice commands because the developer changed the menu structure and didn't document it."
• "I bought a head tracker, a sip-and-puff device, and two different voice recognition subscriptions trying to get one game to work. It still doesn't fully work."
• "I asked a friend to do the character-creation screen for me because the slider controls don't respond to head movements, and I can't use a mouse anymore."
• "I mapped my head movements to keyboard inputs through three separate pieces of software. It breaks every few weeks and I have to rebuild the whole stack from scratch."
• "I stopped playing that game. The developer never responded when I reported the input accessibility problem, and I couldn't find a workaround that didn't lag too much to be playable."
• "I used the system voice control but it doesn't understand my speech pattern well enough for fast-paced commands. I missed too many cues."
• "My bar exam was almost derailed because my usual AT gets flagged by proctoring software. I had to find out whether the exam board would accept a different tool, file the accommodation request, get it approved, and test it, all while studying."

---

All observations occur within the context of accessible computing and gaming technology in the United States.

Footnotes

1. CPARF, "Innovation Insights: Cephable," November 15, 2023. https://cparf.org/innovation-insights-cephable/ ↩ ↩² ↩³

2. Jeffrey Howard, "How AI Is Improving Accessibility for Professionals With Disabilities," InclusionHub, May 22, 2025. https://www.inclusionhub.com/articles/how-ai-is-improving-accessibility-for-professionals-with-disabilities ↩

3. CPARF, "Innovation Insights: Cephable," November 15, 2023. https://cparf.org/innovation-insights-cephable/ ↩ ↩² ↩³

4. Josh Anderson, "ATU706 – Cephable with Alex Dunn," Easterseals Technology, December 6, 2024. https://www.eastersealstech.com/2024/12/06/atu706-cephable-with-alex-dunn/ ↩ ↩² ↩³

5. R.G. Maling and D.C. Clarkson, "Electronic controls for the tetraplegic (Possum) (Patient Operated Selector Mechanisms, P.O.S.M.)," Paraplegia (now Spinal Cord), 1963. https://pubmed.ncbi.nlm.nih.gov/14113294/ ↩

6. ADA National Network, "Digital Access and Title III of the ADA." https://adata.org/research_brief/digital-access-and-title-iii-ada ↩

7. Section508.gov, "Section 508 Law." https://www.section508.gov/manage/laws-and-policies/section-508-law/ ↩

8. Wikipedia, "Game accessibility." https://en.wikipedia.org/wiki/Game_accessibility; IGDA Game Accessibility Special Interest Group, https://igda-gasig.org/ ↩ ↩²

9. Wikipedia, "AbleGamers." https://en.wikipedia.org/wiki/AbleGamers ↩

10. Marketing Charts / PopCap Games, "Casual Gaming and Disability," 2010. https://www.marketingcharts.com/industries/media-and-entertainment-4920 ↩

11. Paul Cairns, Christopher Power, Mark Barlet, Gareth Haynes, Christopher Kaufman, and J. Beeston, "Enabled Players: The Value of Accessible Digital Games," Games and Culture, 2021. https://journals.sagepub.com/doi/abs/10.1177/1555412019893877 ↩

12. Characteristics and Motivations of Players with Disabilities, arXiv preprint. https://arxiv.org/pdf/1805.11352 ↩

13. ALS Forums, Tobii Dynavox TD I-16 pricing, https://www.alsforums.com/community/threads/tobii-dynavox-td-i-16-eyegaze-machine-with-stand-new-20k-new.54423/; Dragon Professional pricing, https://www.thesmbguide.com/dragon-by-nuance ↩

14. National Council on Disability, "Federal Policy Barriers to Assistive Technology," 2000. https://www.ncd.gov/assets/uploads/reports/2000/ncd-federal-barriers-assistive-tech-2000.pdf; Institute of Medicine, "Coverage of Assistive Technologies and Personal Assistive Services," in The Future of Disability in America, National Academies Press, 2007. https://www.ncbi.nlm.nih.gov/books/NBK11441/ ↩

15. National Disability Institute, "The Extra Costs of Living with a Disability in the U.S.," October 12, 2020. https://www.nationaldisabilityinstitute.org/reports/extra-costs-living-with-disability/ ↩

16. WHO and UNICEF, Global Report on Assistive Technology, May 2022. https://www.who.int/publications/i/item/9789240049451 ↩

17. Marta Russell, Beyond Ramps: Disability at the End of the Social Contract (Monroe, ME: Common Courage Press, 1998); Marta Russell, ed. Keith Rosenthal, Capitalism and Disability: Selected Writings by Marta Russell (Chicago: Haymarket Books, 2019). https://www.haymarketbooks.org/books/1289-capitalism-and-disability ↩

18. Alison Kafer, Feminist, Queer, Crip (Bloomington: Indiana University Press, 2013), p. 27. https://iupress.org/9780253009340/feminist-queer-crip/ ↩

19. Katie Ellis, Tama Leaver, and Mike Kent, eds., Gaming Disability: Disability Perspectives on Contemporary Video Games (London: Routledge, 2023). https://www.routledge.com/Gaming-Disability-Disability-Perspectives-on-Contemporary-Video-Games/Ellis-Leaver-Kent/p/book/9781032372853 ↩

20. Ana Londral, "Assistive Technologies for Communication Empower Patients With ALS to Generate and Self-Report Health Data," Frontiers in Neurology, April 26, 2022. https://pmc.ncbi.nlm.nih.gov/articles/PMC9090469/ ↩

21. "Playing online videogames -- more than just entertainment? A qualitative study of virtual social participation in persons with spinal cord injury," PMC, 2024. https://pmc.ncbi.nlm.nih.gov/articles/PMC11137226/ ↩

22. MGMT Boston, "Cephable," December 2024. https://mgmtboston.com/cephable/ ↩

23. Cephable, "Cephable's SXSW Pitch 2024 Victory," March 18, 2024. https://cephable.com/2024/03/18/cephables-sxsw-pitch-2024-victory-riding-the-innovation-wave/ ↩

24. Cephable, "SDK Trial," developer testimonials. https://cephable.com/sdk-trial/ ↩

25. MGMT Boston, "Cephable," December 2024. https://mgmtboston.com/cephable/ ↩