Technology

Enterprises recalibrate strategies to overcome challenges in scaling Generative AI

Friday, 14 November 2025 5:05AM UTC

After initial excitement, large organisations face a critical 'Trough of Disillusionment' with Generative AI, prompting a strategic shift towards responsible, value-driven implementation amidst technical and organisational hurdles.

For nearly three years, Generative AI (GenAI) has stirred significant excitement across global enterprises, promising transformative impacts on customer experience, productivity, and revenue generation. Yet, many large organisations now find themselves amidst a reality check, with market analysts noting that GenAI has entered the so-called "Trough of Disillusionment." This phase reflects a growing awareness of the technology's true capabilities and inherent limitations after initial exuberance.

The journey of GenAI adoption reveals a complex landscape of challenges for big businesses. Chief among these are the gaps between heavy investment and tangible returns, poor data quality, insufficient risk controls, and often unclear business value. Organizational readiness emerges as a critical bottleneck: many companies lack the foundational data infrastructure and AI literacy needed to scale GenAI effectively. Early-stage organisations struggle with identifying viable use cases and tend to harbour unrealistic expectations, while more mature firms confront talent shortages and the need to foster AI literacy across teams. These difficulties are compounded by ongoing struggles with data quality, as learning models inevitably depend on the integrity of their training data, poor data yields unreliable outcomes.

Further complicating rollouts are governance and compliance concerns. Early adopters report challenges such as model hallucinations, bias in AI outputs, and compliance with emerging regulations like the EU AI Act, which sets stringent legal standards for AI use. These insights underscore that GenAI deployment is not merely a technical endeavour but a multifaceted challenge involving people, processes, and organisational culture. Isolated innovation efforts often falter without broad cross-functional engagement aligned to clear business goals.

In response, enterprises seeking to avoid project failure and realise genuine value from GenAI are recommended to adopt a strategic, value-driven approach. This begins with clearly defined business cases and success metrics, focusing on high-impact use cases where GenAI can address concrete problems, such as automating costly tasks or enhancing customer service efficiency, rather than pursuing AI for AI's sake. A rigorous upfront analysis of costs versus potential benefits is also essential.

Cross-functional collaboration is another pillar of successful adoption. Bridging the traditional divides between IT, data science, business units, and risk management ensures a deeper understanding of both business context and technological constraints. Establishing governance structures like AI councils can help align initiatives with enterprise strategy while safeguarding ethical and compliance standards. Cultural readiness must not be neglected, with change management and upskilling programmes necessary to help employees embrace and effectively use AI tools. Pilot projects involving end users have proven beneficial in gaining feedback and building incremental success, which is crucial to maintaining momentum and managing expectations.

Experts also advise using a structured operating model that transitions AI adoption from ideation through to industrial-scale impact. This could involve a three-phase framework: discovery and baselining, which assesses current readiness and defines priority use cases; tooling and design, focused on building scalable, secure, and governed solutions integrated with business workflows; and finally, ROI and scaling, where value is proven in limited deployments before broader institutionalisation across the enterprise. Integration of Responsible AI practices, including bias detection, safety testing, human oversight, and continuous monitoring, is essential throughout all phases to maintain trust and accountability.

Real-world success stories illustrate this approach. An Australian bank, for example, applied GenAI to expedite its software testing process, traditionally a slow and manual affair. This led to accelerated release cycles and improved software quality, enhancing customer outcomes and fostering a culture of collaboration. In the pharmaceutical sector, a North American company leveraged GenAI to transform compliance and audit operations, slashing manual document review times by 65% and detecting potential quality gaps with over 95% accuracy.

Nevertheless, scaling GenAI remains a marathon rather than a sprint. While many enterprises now face the trough of disillusionment where initial deployments have not met lofty expectations, this phase is survivable with recalibrated strategies. Addressing fundamental issues like data quality, organisational readiness, and fostering collaboration across business and technical teams are crucial to avoiding common pitfalls. Setting realistic milestones and embracing a product-aligned operating model focused on measurable business outcomes can unlock sustained AI value.

Additional industry perspectives reinforce these findings. Challenges such as a lack of measurable objectives, data and knowledge management hurdles, shortage of AI expertise, and organisational resistance routinely impede AI scaling efforts. Experts recommend establishing AI Centres of Excellence, robust AI strategies, and investing in data infrastructure and literacy to overcome these barriers. Ensuring strong executive sponsorship and embedding responsible AI governance are also pinpointed as critical for success. Security and compliance concerns, especially data privacy risks in sensitive sectors, necessitate rigorous controls including encryption, access management, and compliance with regulations like GDPR or HIPAA.

Emerging best practices include standardising AI tools and processes, creating centralized knowledge repositories, and streamlining organisational access to advanced technologies. Managing costs, especially around customisation and operational overheads, is another practical consideration. Above all, enterprises must position GenAI deployment as a holistic transformation that aligns technology advancements with people, processes, and strategic purpose.

In sum, although Generative AI’s promise continues to captivate, delivering on that promise requires disciplined, strategic action grounded in realistic understanding and responsible governance. Enterprises already embracing this comprehensive approach are beginning to convert early investments into sustainable, measurable business value.

📌 Reference Map:

^[1] (TechRadar Pro) - Paragraphs 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12
^[2] (Inbenta) - Paragraphs 3, 4, 6
^[3] (Shieldbase) - Paragraphs 3, 4, 6
^[4] (XenonStack) - Paragraphs 2, 7
^[5] (Deloitte Insights) - Paragraph 3
^[6] (Oyelabs) - Paragraphs 2, 7
^[7] (Teneo AI) - Paragraph 7

Source: Noah Wire Services

More on this

https://www.techradar.com/pro/how-big-businesses-are-handling-the-roll-out-of-generative-ai - Please view link - unable to able to access data
https://www.inbenta.com/articles/5-challenges-businesses-face-in-adopting-ai-at-scale/ - This article discusses five key challenges businesses encounter when scaling AI adoption: lack of measurable business objectives, data and knowledge management issues, shortage of AI expertise, choosing the wrong AI partner, and organizational resistance to change. It emphasizes the importance of a strategic approach, including establishing AI Centers of Excellence, developing comprehensive AI strategies, investing in data infrastructure, fostering AI literacy, and focusing on continuous improvement to overcome these obstacles and unlock AI's full potential.
https://shieldbase.ai/blog/top-challenges-enterprises-face-when-scaling-ai - This blog post outlines several challenges enterprises face when scaling AI, such as aligning AI initiatives with measurable business outcomes, securing executive sponsorship, overcoming organizational resistance to change, addressing data privacy and security concerns, managing technical and operational challenges, scaling AI responsibly, and measuring ROI. It suggests that a clear roadmap linking AI initiatives to business outcomes, strong executive support, and responsible AI practices are essential for successful AI scaling.
https://www.xenonstack.com/blog/challenges-opportunities-generative-ai - This article highlights key challenges in adopting agentic AI, including governance and compliance, balancing costs with ROI, integration complexity, and trust and ethical risks. It emphasizes the need for responsible AI frameworks, bias detection, explainability standards, and continuous monitoring to address these challenges effectively and build trust in AI systems.
https://www2.deloitte.com/us/en/insights/industry/public-sector/government-faces-challenges-with-generative-ai-adoption.html - This Deloitte Insights report examines the challenges government organizations face in scaling generative AI, such as lack of expertise and difficulty measuring mission value. It notes that government respondents report the least expertise in generative AI and are the least prepared from a talent perspective, which explains why choosing the right technologies is the biggest reported barrier for the public sector adopting generative AI.
https://oyelabs.com/challenges-in-integrating-generative-ai-into-business/ - This article discusses challenges in integrating generative AI into business, including data privacy and security risks, high implementation and operational costs, and the need for substantial financial investment in infrastructure, computing power, and skilled personnel. It emphasizes the importance of ensuring compliance with regulations such as GDPR, CCPA, and HIPAA to protect user data and implementing robust encryption, access controls, and AI auditing mechanisms to mitigate these risks.
https://www.teneo.ai/blog/scaling-generative-ai-5-conversational-ai-challenges-solutions - This blog post identifies five challenges in scaling generative AI: context management, lack of domain-specific knowledge, scalability and latency issues, compliance and security concerns, and cost of customization. It suggests that enterprises should address these challenges by developing standardized tools and predefined patterns for common generative AI applications, creating a centralized repository for well-documented patterns, code templates, and architecture diagrams, and facilitating organization-wide access to advanced tools through streamlined processes for requesting access, training, and onboarding.

Noah Fact Check Pro

The draft above was created using the information available at the time the story first emerged. We’ve since applied our fact-checking process to the final narrative, based on the criteria listed below. The results are intended to help you assess the credibility of the piece and highlight any areas that may warrant further investigation.

Freshness check

Score: 8

Notes: The narrative presents recent insights into Generative AI adoption challenges, with references to Gartner's 2025 Hype Cycle and other current sources. ([procurementmag.com](https://procurementmag.com/news/gartner-generative-ai-trough-disillusionment?utm_source=openai)) However, similar discussions have been published in the past year, indicating some recycled content. ([cio.com](https://www.cio.com/article/4046443/gen-ai-descends-into-disillusionment.html?utm_source=openai))

Quotes check

Score: 7

Notes: The report includes direct quotes from Gartner analysts and other experts. ([procurementmag.com](https://procurementmag.com/news/gartner-generative-ai-trough-disillusionment?utm_source=openai)) While these quotes are relevant, they have appeared in previous publications, suggesting potential reuse. ([cio.com](https://www.cio.com/article/4046443/gen-ai-descends-into-disillusionment.html?utm_source=openai))

Source reliability

Score: 9

Notes: The narrative originates from TechRadar Pro, a reputable technology news outlet. The inclusion of references to Gartner's 2025 Hype Cycle and other credible sources enhances its reliability. ([procurementmag.com](https://procurementmag.com/news/gartner-generative-ai-trough-disillusionment?utm_source=openai))

Plausability check

Score: 8

Notes: The challenges outlined, such as data quality issues, insufficient risk controls, and unclear business value, are consistent with known obstacles in Generative AI adoption. ([cio.com](https://www.cio.com/article/4046443/gen-ai-descends-into-disillusionment.html?utm_source=openai)) However, the narrative's tone is somewhat dramatic, which may warrant further scrutiny.

Overall assessment

Verdict (FAIL, OPEN, PASS): FAIL

Confidence (LOW, MEDIUM, HIGH): MEDIUM

Summary: The narrative presents valid concerns regarding Generative AI adoption challenges, supported by reputable sources. However, the recycling of content and quotes from previous publications, along with a somewhat dramatic tone, raise questions about its originality and objectivity. ([cio.com](https://www.cio.com/article/4046443/gen-ai-descends-into-disillusionment.html?utm_source=openai))

Generative AI
AI adoption
Enterprise transformation