BLOG
From hurdles to breakthroughs with AI
Lessons from leadership discussions hosted by Accenture & AWS
10-MINUTE READ
November 24, 2025
In October and November of 2025, Accenture convened a select group of joint AWS clients for a series of intimate, cross-industry discussions on the realities of AI adoption in their organizations. Among the business leaders attending were top decision-makers ranging from COOs to CIOs, CDOs, Heads of Strategy and Chief Product Officers. The depth and candor of these sessions surfaced a set of shared challenges and nuanced responses—resulting in insights that offer a unique lens on the evolving landscape of AI in business.
Let’s explore what leaders are saying about how they’re approaching some persistent hurdles to shape and scale AI adoption today:
Leaders are redefining ROI
ROI has been difficult to pin down, but leaders are looking beyond just productivity and revenue gains.
The challenge
Leaders tend to focus on AI business cases with clear savings or major productivity outcomes. But ROI can be harder to quantify for other projects with smaller efficiency gains (like email automation) or in hard-to-quantify areas like knowledge management, collaboration and coaching. Leaders in low-cost labor markets also have trouble justifying ROI using conventional metrics. Low-risk projects are still widely favored, particularly in regulated industries.
What leaders are saying
Leaders are expanding their view of ROI beyond traditional metrics—some questioning if they’re even relevant. They see value in outcomes like creating engaging employee experiences and preparing teams for the future of work. One participant found that enthusiasm for AI in their organization is driving digital transformation efforts, which generates immediate and far-reaching value.
How to take action
Organizations that think big when it comes to AI and align technology goals with business and talent objectives tend to achieve more value.
- Target more holistic outcomes
Tackle core business challenges with measurable, 360° value outcomes that are both financial and non-financial. For example, workflow redesign may not render productivity gains overnight, but this work yields rewards that will continue to resonate across people and the business. Also, operationalizing responsible AI delivers value across finances, experience, talent, sustainability, inclusion & diversity and risk mitigation.
- Lead with value
Successful leaders know they’re in it for the long haul. Prioritize AI investments that drive long-term innovation and growth, not just short-term cost savings and productivity. The C-suite must set clear priorities and benchmarks for the company’s AI investments to move through true transformation confidently.
- Map KPIs to target value
Each AI use case will have unique value to the business. Align KPIs to these strategic objectives rather than just implementation targets. The business must provide structured capital allocation to support such key initiatives. Track results continuously to make sure AI delivers sustained business value.
34%
of organizations have scaled at least one industry-tailored solution for a core process.
3X
These companies were 3X more likely to have achieved better than expected ROI.
18%
average AI revenue growth estimated for companies with mature responsible AI.
AI opens doors to tackle longstanding modernization challenges
Eagerness to unlock AI value is driving leaders to address lingering readiness gaps by optimizing their technology and operations.
The challenge
Technical roadblocks are impeding leaders on their journey to realize all AI has to offer. Some top challenges include 1) Data readiness: Siloed data is the scourge of AI value. Leaders cite difficulty of migrating data to the cloud, security issues and complex governance and documentations. Lack of enterprise-wide data classification is also holding them back. 2) Trust: Leaders face complexities related to the security and safety of the models as well as questions around reliability. Some even cite challenges with models handling certain languages and dialects. 3) Legacy infrastructure: Outdated infrastructure makes it harder to unify and access data, enforce security and governance, and power modern, integrated AI ecosystems.
What leaders are saying
Leaders are approaching digital transformation with renewed urgency: They’re modernizing legacy systems, standardizing tools and platforms, and focusing on improving data quality and governance to unlock AI value. One participant even noted that their company has reduced legacy systems by over 50% to streamline data management and integration.
How to take action
Without modern technology foundations, optimized processes and AI-ready data, AI implementations will struggle. The good news? AI is a catalyst—and a critical asset—for accelerating end-to-end modernization.
- Unlock data to power AI
To prepare data for AI, establish a company-wide framework to understand your data’s value and prioritize optimization. Redesigning data architectures with comprehensive governance is also essential. Use AI tools to boost data readiness—for example, by accelerating data migration and modernization or automating security and compliance.
- Prioritize AI ecosystem interoperability
With the AI ecosystem expanding, choice paralysis is a common roadblock. Develop a strong ecosystem strategy to guide current and future implementations. Dedicate investments to building interoperability across your AI systems and providers to support scalability and avoid fragmentation.
This approach supports compliance and resilience across data, infrastructure, models, agents and apps. It also enables the development of localized AI that’s tailored to languages, cultures and priorities—unlocking new value opportunities.
- Build organizational and technical agility
Create a flexible, scalable infrastructure, taking advantage of omni-cloud offerings. Adopt platform approaches to ensure reusability and consistency in how AI is deployed. Embrace modern engineering practices such as reusable code, agile methodologies and FinOps to keep teams executing efficiently and make sure AI initiatives deliver continuous value.
Integrating sovereign AI solutions can help address security and compliance complexities, especially for business-critical or highly sensitive use cases. This approach supports compliance and resilience across data, infrastructure, models, agents and apps. It also enables the development of localized AI that’s tailored to languages, cultures and priorities-unlocking new value opportunities.
Just 13%
of companies are “extremely confident” they have the data strategies and digital capabilities for AI.
48%
of organizations lack enough high-quality data to operationalize their gen AI initiatives.
60%
of C-suite leaders are prioritizing investments in strengthening their digital core.
Visionary leadership is needed to address deep-rooted friction
Roadblocks across people and processes slow AI progress, but committed leadership and new governance models yield breakthroughs.
The challenge
Organizations are not always designed to welcome rapid change—with good reason. But that can come at odds with the kind of transformation AI requires. Participants introduced the idea of “organizational antibodies”: deeply embedded systems and processes that are designed to protect the business—like regulatory rigor and security audits—but that can cause friction if not actively addressed.
What leaders are saying
Strong leadership is needed to find pathways through friction. Participants say success hinges on not just securing early buy-in but actively engaging leadership and business stakeholders, and building flexibility into the organization. Some companies are engaging middle managers to evangelize change and secure alignment, and empowering domain experts to guide AI use case development and governance. This marks a shift toward a hub-and-spoke model, which leaders see as a “game changer” for AI progress.
How to take action
Navigating transformation to achieve AI’s full potential requires new forms of leadership and more flexible ways of operating.
- Actively engage leadership
Digital fluency drives reinvention. Help boards, executives and managers develop a deeper understanding of the technology, along with its business impact, to enable faster decisions and bolder strategies. Leadership should set the vision, secure resources and align the organization, embracing the core human principles of curiosity, courage and connection.
- Evolve toward less-centralized AI governance
Centralized Centers of Excellence (CoE) should shift toward new Capability Building Teams (CBT), responsible for building multi-disciplinary cohorts focused on implementing AI for specific challenges. CBTs can also distribute insights and best practices by embedding them into widely available AI agents.
- Collaborate across disciplines
As AI reshapes traditional hierarchies and boundaries, cross-functional collaboration is key to driving success. Use AI to quickly distill and share insights and information across teams. Breaking down silos allows AI, data, business and governance teams to work in sync to accelerate impact.
6X
greater likelihood of achieving business value when leaders deeply understand gen AI.
65%
of executives say they lack the expertise to lead gen AI transformations.
97%
of executives believe gen AI will transform their companies and industries.
AI leaders seek to make work better for workers
Employee fears are real, but leaders are looking to evolve work—and ways of working—rather than replace people.
The challenge
Leaders cite persistent fears among employees that AI may take their jobs, particularly for organizations early in their adoption journeys. To allay these fears, organizations acknowledge that they need to help employees understand how AI can enhance their individual roles and how to collaborate with it effectively.
What leaders are saying
Participants who surfaced employee fears about AI also indicated their organizations are not looking for 1:1 replacement. Instead, they’re focused on transforming work through human and AI partnership. They acknowledge a need for ongoing reskilling and learning for both IT and business teams, including education on effective AI interaction and prompting. Commitment to people and trust in technology continue to shape the conversation.
How to take action
Leaders must put human wellbeing and creativity at the center of AI-driven reinvention to unlock opportunities for more meaningful, impactful work.
- Evolve talent models to empower people
As work evolves, employees gain the flexibility to foster new competencies and move within a newly fluid organization. This enables them to build diverse skills and experiences, boosting their relevance in the market. Organizations need agile talent models and a culture where employees can actively shape their work and its impact across the company.
- Invest in learning and preparedness
Comprehensive, continuous learning initiatives are vital to equip people with the skills and capabilities needed to collaborate with machines. Companies must build a strong teach-to-learn culture that approaches learning from three perspectives: individual, organizational and the machine itself.
- Shift to human + machine co-learning
Traditional training paradigms are being rewired: Continuous co-learning emphasizes dynamic collaboration between people and AI. To facilitate this, leaders must promote curiosity, integrate learning into daily work, build transparency and accountability, and provide intuitive AI tools with the right coaching and support.
2/3
of Reinventors strongly agree that work will become more meaningful, creative and impactful with gen AI.
3X
more gen AI budgets are spent on technology than on people.
84%
of executives expect regular human-AI collaboration within 3 years
Process reinvention is key to value
With mounting pressure to apply AI everywhere, leaders are first taking a hard look at their processes to ensure value.
The challenge
Leaders feel pressure to expand AI across the business, but doing “AI for AI’s sake” can lead to superficial or misaligned implementations. One participant described AI as “a very expensive band-aid to a bad process,” illustrating the risk of applying AI to already-flawed workflows.
What leaders are saying
Leading organizations are focusing on process reinvention, questioning whether certain processes are even necessary and how they can be augmented or transformed with AI. This sharp focus is a key driver of effectiveness.
How to take action
To get the most out of AI, leaders need to revisit entire workflows, individual processes and organizational structures—keeping both human and AI potential top of mind.
- Rethink workflows
Reimagine workflows to identify where AI drives the most impact. Align technology with business goals, strategically break down silos and reallocate work to better serve customers, people and business goals.
- Redesign the organization
As jobs change, the human + AI partnership will become a core component of organizational structures. Facilitate this shift by breaking down silos, making data accessible, introducing multi-agent capabilities and fostering dynamic skills to unlock this partnership’s full potential.
- Reinvent processes with AI
AI is a powerful tool to reinvent entire domains and processes. Use AI to “hack” workflows by rooting out inefficiencies, minimizing manual steps and uncovering opportunities missed due to biases or siloed thinking.
52%
of Reinventors are reshaping the workforce by redesigning jobs and roles around gen AI.
47%
recognize that their processes will require significant change.
75%
are reskilling people and actively involving them in enterprise change efforts.
Reimagining success: Bold moves build a human-centered AI future
These candid discussions with our joint AWS clients reflect the current reality of AI adoption: Organizations are eager to realize the promise of AI, but still find efforts tied up by technical, cultural and strategic challenges. However, leaders are already embracing new strategies to conquer this complexity. Those poised to succeed are willing to rethink ROI, tackle technical readiness and organizational readiness, build strong leadership and reinvent processes for human + AI collaboration—all while prioritizing and supporting their people through change.
Special acknowledgement to the co-authors of this blog and co-facilitators of this discussion series: Ariel Berstein, Satish Lakshmanan, and Pavan Sethi. Thank you also to Jennifer Jackson and Allister Fraser, Global Leads for the Accenture AWS Business Group.
Want to learn more or schedule a workshop? Get in touch.