Why do some organizations fail to see ROI from AI and how can they correct their course in the AI journey?
MH: Another common mistake is that AI projects often lack clear business alignment. Many companies experiment with AI but fail to tie it to specific KPIs or operational goals. This is why we see a disconnect; 42% of enterprises in Asia consider AI critical to their business, yet their AI models produce accurate results only 32% of the time. That’s a sign that while AI adoption is high, execution is still falling short.
A lot of companies jump into AI expecting instant results, but the reality is that without the right groundwork, ROI is difficult to achieve. One of the biggest reasons for this is poor data quality and availability.
Our research shows that, in Asia, data is only available when needed 34% of the time, which means AI models often lack timely and accurate input. Even when businesses invest in AI, the technology can’t deliver real value if it’s working with incomplete, inconsistent, or low-quality data.
So how do companies fix this? First, AI initiatives need to be aligned with real business needs, not just pursued for innovation’s sake. Second, organizations need to prioritize data readiness before scaling AI, ensuring that models are trained on high-quality, well-structured data.
Finally, AI projects must be treated as long-term investments, with the right MLOps, governance, and performance monitoring in place. The businesses that are seeing real ROI aren’t just experimenting with AI – they’re integrating it into their workflows with clear goals, robust data strategies, and continuous optimization.
What are some key considerations for AI governance and compliance amid evolving regulations?
MH: As AI adoption picks up speed, governance is becoming a bigger concern especially when it comes to data security and compliance. In Asia, 44% of enterprises see data security as a top issue, which is even higher than the global average of 38%. The challenge is even more pressing in markets like India (54%) and Indonesia (50%), where stricter regulations and data sovereignty laws are raising the stakes for AI deployment.
For AI to be truly effective and trustworthy, organizations need to ensure their models are fair, transparent, and explainable. AI-driven decisions should be auditable so businesses can maintain accountability, and ongoing monitoring is key to preventing bias. Without proper governance, AI can easily become unpredictable or worse, create compliance headaches.
Another big piece of the puzzle is data protection and regulatory compliance. With AI making more critical decisions than ever, companies must navigate increasingly complex data privacy laws while ensuring their systems remain secure, especially when dealing with cross-border data flows.
Governance is a defining factor in AI success, with 45% of top-performing AI adopters in Asia crediting strong governance frameworks as a key reason for their achievements – well above the global average.
The best approach is to bake governance into AI strategy from the start, rather than treating it as an afterthought. Organizations that invest in automated governance tools and proactive compliance measures will be in a much better position to keep up with evolving regulations without slowing down innovation.
In your opinion, what is the next wave of AI-driven automation in software, cloud, and enterprise workflows?
MH: AI is shifting from being a tool for experimentation to a driver of enterprise automation, and the next big wave is all about self-optimizing systems. We’re seeing major advancements in AI-powered software development, where generative AI and automation tools are helping developers write, test, and deploy code faster than ever. This isn’t about replacing developers but about accelerating software delivery and reducing costs.
Beyond software, AI-driven IT operations (AIOps) are changing the way businesses manage cloud infrastructure. AI can now predict system failures, optimize cloud costs, and automate maintenance, reducing downtime and improving efficiency. With data volumes soaring and cloud costs becoming a major concern, enterprises will need AI to keep IT environments running smoothly.
AI is also making enterprise workflows smarter. We’re moving beyond basic robotic process automation (RPA) into cognitive automation, where AI can handle complex decision-making in areas like finance, HR, and supply chain management. This means AI isn’t just automating repetitive tasks – its helping businesses make faster, more informed decisions.
One of the most exciting developments is AI at the edge. With AI models running closer to where data is generated whether in factories, hospitals, or smart cities, businesses can analyze and act on data in real time. This reduces latency and enables faster, more responsive automation across industries.
39% of IT leaders in Asia now rely on third-party AI expertise, signaling a shift toward vendor-supported AI solutions rather than purely in-house development. This trend highlights the growing need for scalable, well-integrated AI automation frameworks that allow enterprises to stay agile while keeping governance and security in check. The companies that get this right will be the ones leading the next wave of AI-driven transformation.
