For years, accessibility was treated as a last‑mile checkbox—something teams tested right before launching, or only after users complained. In a world reshaped by AI, that model is no longer acceptable. Accessibility now must be continuous, intelligent and embedded into every stage of the digital journey, from design to production and ongoing support.

From compliance to culture: HCLTech’s accessibility journey

HCLTech’s partnership with Microsoft on accessibility spans more than a decade and the story has evolved from narrow compliance to deeply embedded cultural change. Early work focused on doing “compliance with good intentions”—learning through trial, error and missteps as the team figured out what it meant to build accessible products and services at scale.

Today, accessibility is woven into how HCLTech operates and how it shows up for employees, customers and communities. Programs include:

• Annual support for visually impaired students with assistive devices to help mainstream them into education
• Device-focused labs where teams test with assistive technologies and advance inclusive design roadmaps
• Recruiting and community outreach that has influenced more than 15,000 people, including employees and community members with disabilities
• Global Accessibility Awareness Day events and immersive training sessions that bring teams closer to real barriers and drive honest conversations

Crucially, people with disabilities are engaged at the very start of design rather than only at testing. HCLTech involves employees with disabilities in “stage one” discovery, valuing their lived experience as a core design asset, not an afterthought. Stories like that of Swati Sena—who joined as an accessibility tester, earned certifications and awards and now leads complex customer experiences—illustrate how inclusive careers are being built inside the company. 

Accessibility as a continuous, AI‑powered lifecycle

During a session at the 2026 Microsoft Ability Summit which featured Carol Criner, Senior Vice President, Strategic Accounts at HCLTech, a simple but profound shift was noted: AI is turning accessibility into a proactive, lifecycle discipline.

Across the software lifecycle, AI is reshaping what’s possible:

• Design: AI can flag contrast issues, focus gaps and usability barriers before a single line of code is written, allowing teams to prevent issues rather than patch them later
• Development: AI detects WCAG violations automatically in delivery pipelines, helping developers ship accessible experiences by default
• Testing: AI accelerates defect detection, generates test scenarios, automates checks and standardizes bug logging, dramatically improving consistency and coverage
• Remediation: AI maps defects to standards, code examples and best-practice fixes, guiding teams to faster and higher‑quality remediation
• Post‑production: AI monitors releases, predicts regressions and connects real user feedback back into accessibility signals to create a continuous improvement loop

Behind this approach is a decade-plus of scaled accessibility delivery for Microsoft. The program has logged and resolved thousands of critical issues and delivered millions in “value ideas” implemented for Microsoft.

HCLTech now delivers end‑to‑end accessibility services—from compliance testing and remediation to usability validation, design assessments, audits, empathy sessions and disability support desks—modernized with AI‑powered tools and global partnerships with people with disabilities.

AI in accessibility and accessibility in AI

Apoorv Iyer, HCLTech’s Global Head of Generative AI, framed the opportunity around a crucial distinction: AI and accessibility versus accessibility in AI. AI and accessibility are the familiar narratives of using AI to improve assistive technologies and accessible experiences. Accessibility in AI asks a more foundational question: are the AI systems themselves inclusive when they mimic and automate human behavior?

Iyer cited research on computer‑use agents that can navigate websites, log in and complete tasks on behalf of users. These systems are often optimized purely for task completion, without considering the accessibility needs of different users or whether the paths they choose are themselves inclusive. That gap shows why accessibility must be built into AI models, tools and interaction patterns, not just layered on top.

Within HCLTech, AI is now infused across multiple layers:

• AI Force: A platform for Agentic AI that is audited against accessibility standards and made usable for employees across roles
• AI Foundry: Data modernization and inclusive data practices that ensure AI models are trained on representative, equitable datasets
• AI Engineering: Examples like “Suno,” a cochlear implant device developed by HCLTech, use AI for simulation, outcome prediction and testing to improve hearing outcomes for users.
• AI Factory & Advisory: Enterprise AI advisory that explicitly considers how accessible users will use AI, ensuring that AI adoption strategies include accessibility as a first‑class requirement.

This dual focus—using AI to expand accessibility and ensuring AI itself is accessible—echoes the broader message of the session: build around human potential, not limitations.

The real shift: Time back, not just barriers removed

Perhaps the most powerful segment of the session came from Ajay Sharma, Accessibility SME and Technical Manager at HCLTech, who is also a screen reader user. He framed the everyday reality for many employees with disabilities: everyone has the same 24 hours, but inaccessible or inefficient experiences can stretch simple tasks from five minutes to fifteen.

Scheduling a meeting, editing a document or understanding a complex UI layout can be almost effortless for sighted users, yet repetitive and mentally draining for screen reader users. Sharma openly shared that he often avoids scheduling meetings himself due to the friction of navigating complex calendar grids and availability views.

Copilot is transforming these experiences:

• In Outlook, Sharma can now simply ask Copilot to “find the earliest available 30‑minute slot” with specific colleagues. Copilot analyzes calendars, presents accessible options and sends the invite—turning minutes of navigation into seconds
• In Word, Copilot becomes an “agentic authoring” assistant, scanning for formatting inconsistencies, fixing headings and spacing and applying consistent, accessible styles based on a single prompt
• With Copilot Vision, Sharma can share his screen and ask what’s there, receive a description of the layout and content, then ask follow‑up questions about charts, infographics or UI structure in natural language

These scenarios showcase AI as an intelligent assistive technology, offering spatial awareness, context and interpretation that traditional screen readers struggle to convey. For Sharma, the true shift is that accessibility is no longer just about removing barriers; it is about giving time back, boosting independence and allowing people to focus on their work instead of constant workarounds.

Two patterns emerge from these examples: a solid, accessible platform and AI layered on top. When the underlying platform (for example, Office canvas, screen readers, assistive technologies) is strong and AI is thoughtfully layered, the result is what he calls “magical.”

Scaling accessibility: People, platforms and mindset

Panelists from the session, including Crystal Jones, Senior Support Delivery Leader, Disability Answer Desk at Microsoft and Clint Covington, Principal Group Manager at Microsoft, expanded the conversation to what it takes to scale accessibility globally.

Several themes surfaced:

• Support as a front line for AI innovation: Disability Answer Desk has experimented with tools like Be My Eyes and AI capabilities that cut resolution time, helping customers solve issues faster and with more independence. HCLTech partnered with Microsoft on “AskMa,” an AI‑powered assistant that lets support staff query internal knowledge for quick answers, such as keyboard shortcuts in Teams.
• Accelerating feature development: AI‑assisted coding and prototyping are enabling accessibility features to be built in weeks rather than months, as in the example of a new LaTeX conversion pipeline that would previously have taken much longer with lower quality.
• Bridging organizational incentives: Covington pointed out a core challenge: the priorities of product leaders (ship on time, grow the business) don’t always naturally align with the needs of the disability community. Addressing this “principal–agent” gap requires education, storytelling and partners like HCLTech who can reduce friction and bring accessibility expertise into product teams.

Mindset shifts emerged as a critical lever:

• Start from the assumption that progress is possible, even if it cannot all be done at once
• Treat compliance as the floor, not the ceiling and reimagine accessibility as expanding human possibility
• Invest in strengthening foundational platforms—screen readers, documents, canvases—so that AI “frosting” can truly deliver magical experiences
• Design AI systems around human potential, not perceived limitations and ensure technology adapts to individual preferences rather than forcing people to adapt to technology

As Sharma stated in his closing remarks, the goal is to look beyond minimum standards and ask what can be done “beyond compliance.” With AI maturing and trust frameworks taking shape, organizations now have the tools to scale accessibility in a meaningful and responsible way.

What leaders can do next

The session closed with a shared call to action for organizations navigating the age of AI:

• Embed accessibility from day one in design, engineering and testing, with people with disabilities involved early and often
• Use AI across the lifecycle—to detect and prevent issues, speed remediation and continuously learn from user feedback—while ensuring AI itself is designed for accessibility
• Invest in culture and capability building: immersive training, accessibility academies and career paths for employees with disabilities
• Partner across the ecosystem—vendors, service providers, hyperscalers and communities—to share patterns, tools and expertise

For HCLTech and Microsoft, “built for inclusion, powered by AI” is more than a tagline; it is a joint commitment to creating accessible ecosystems where every individual has the power to thrive.