XChange Asia 2026: AI to squeeze out transactional SIs as Asia channel pivots to outcomes

Channel ecosystem leaders say uneven AI maturity across Asia is forcing partners to specialise, build governance-led offerings and adopt outcome-based pricing to push customers beyond pilots.

A panel discussion held on the sidelines of Xchange Asia in Singapore recently, moderated by CRN Australia Editor Athina Mallis, brought together channel ecosystem leaders from the region. The panel featured Sam Chng, MD, SiS Technologies; Maury G, Global GTM and Strategic Advisor at Pivotale AI; and Ananth Lazarus, MD, GTDC APJ.

Titled “The Channel’s AI Reality Check: Where Partners Win and Stay Relevant”, the discussion focused on how AI adoption is moving beyond pilots and into areas of measurable commercial value, and what that means for system integrators, managed service providers and the broader partner ecosystem.

As AI adoption accelerates across Asia, the leaders said the region remains highly uneven in maturity.

SiS Technologies’ managing director, Sam Chng, said AI adoption across Asia varies sharply by market, with China leading at a significantly higher level of maturity.

“In China, AI adoption is very strong and widely discussed. I would rate it around eight out of ten. North Asia is also relatively advanced, but South Asia is around six out of ten,” he said.

He added that South Asia’s lower rating is largely due to the stage of engagement, where many organisations are still in an educational phase, trying to understand AI’s role in workflows, processes and cybersecurity.

“A lot of organisations are still asking what AI can do, how it can help processes, workflows, cybersecurity, and so on. There is still a significant amount of education required,” he said.

According to Chng, deep AI adoption remains limited in the region despite growing interest.

Traditional channel roles face structural pressure

On the question of how AI is reshaping partner roles, Chng said transactional system integrators are unlikely to survive in their current form.

“Transactional system integrators will be eliminated very soon. If you are only doing transactional work, there is no value to the customer,” he said.

He added that organisations are already being pushed to use AI for troubleshooting and support, which is gradually shifting traditional service models.

Over time, he said level one support functions will also evolve, changing the structure of the services market and impacting existing revenue streams built around troubleshooting, email support and phone-based services.

However, he noted that this shift is also creating new opportunities.

“Everything from workflow to cybersecurity will be AI-powered. That creates opportunities in AI skillsets,” he said.

He also pointed to a broader structural change in the channel ecosystem, where no single company will be able to cover the entire stack.

“Moving forward, I don’t think there will be one company that can do everything. Partnerships will become more important because every partner will have a specific skillset,” he said.

AI expected to reshape channel structure through consolidation and new service models

Pivotale AI’s global GTM, strategic advisor, Maury G, said AI will significantly disrupt the channel ecosystem, leading to consolidation and a shift in traditional specialisations.

AI is also driving the emergence of new categories similar to earlier “as-a-service” models, he said.

“We will see AI-as-a-service models across infrastructure, application layers, and security layers,” he said.

While the disruption will be structural, he added that it will also create new opportunities, particularly for partners who move towards advisory and consultation-led models.

“Those who are not using AI will be left behind. Many will shift towards being strategic advisors, and consultation-driven models will create new opportunities across the channel,” he said.

AI-native partner ecosystem beginning to emerge

GTDC’s managing director for APJ, Ananth Lazarus, said the partner landscape is already undergoing a shift similar to what was seen during the cloud transition, when customer demand began redefining partner roles.

He said new AI-first and agentic-native partners are now emerging, focused on designing, deploying and running AI-driven processes within organisations.

“No single company can be a master of everything, so partnerships will be essential in this new ecosystem,” he said.

Lazarus also pointed to emerging ecosystems forming around players such as Nvidia, Anthropic and OpenAI.

At the same time, Lazarus mentioned that advisory-only roles will gradually lose relevance.

“Advisory-only roles will reduce because companies now want outcomes, not just advisory services. If you cannot deliver outcomes, you will be out,” he said.

AI adoption moving faster than cloud, but still early in Asia

On the slow conversion from AI pilots to production, Lazarus said the transition is expected to be faster than the cloud era, where on-premise to cloud migration was significantly more complex and time-consuming.

However, he noted that AI adoption in Asia is still at an early stage.

“It is still nascent, especially in Asia, where adoption remains quite surface-level. But I am optimistic that adoption will accelerate. It feels like a lull before the storm,” he said.

Maury attributed the hesitation to multiple forms of fear within organisations, including job security concerns, fear of failure, and uncertainty around AI reliability.

“There is fear of losing jobs if AI performs well, fear of failure, and fear related to AI hallucinations and lack of guaranteed accuracy,” he said.

He added that organisations also struggle with the risk of moving from experimental to production environments due to data security concerns and a lack of guaranteed outcomes.

Chng, however, framed the issue differently, calling it caution rather than fear.

“Adoption is slow, but not because people are not interested. It is because they are careful and prefer a staged approach,” he said.

He added that this staged approach requires time and the right skillsets, as AI increasingly converges across networking, security, infrastructure, workflows and operations.

“The challenge is that we do not yet have enough talent or skillsets to manage this convergence,” he said.

Governance, talent and commercial opportunity key to unlocking adoption barriers

On how to reduce fear and caution, Maury said governance and risk frameworks will play a critical role in enabling adoption, alongside organisational reassurance that AI is about augmentation rather than replacement.

He pointed to recent messaging from China that AI will not replace jobs as an example of workforce reassurance.

“Governance and HR need to assure employees that AI is not about replacement but augmentation,” he said.

Lazarus agreed on the importance of governance and cybersecurity as foundational requirements for enterprise AI adoption.

Maury also highlighted a significant commercial opportunity for partners, noting that AI is expanding the services-led value pool across the ecosystem.

With AI, service-led opportunities linked to security and deployment are expected to increase substantially, he said.

Chng added that workforce shifts are already visible, with many organisations hiring younger, AI-trained talent who can adapt more quickly and at lower cost compared to senior executives.

Specialisation, IP and outcome-based models define future partner strategy

On where partners should invest, Chng said the future belongs to specialised AI integrators rather than broad-based system integrators.

“System integrators will evolve into AI integrators, covering workflow, governance, cybersecurity, and network domains powered by AI,” he said.

Maury said partners should focus on a core area while remaining vendor agnostic, given the uncertainty around which AI vendors will remain viable in the coming years.

He cautioned against over-dependence on any single vendor due to rapid ecosystem shifts driven by funding, regulation and market dynamics.

Lazarus said partners must avoid horizontal expansion and instead focus on deep domain expertise in select industries.

He emphasised three priorities. Building IP in two core industries, investing heavily in AI-ready talent, and shifting towards outcome-based pricing models.

Lazarus added that customers today are reluctant to move beyond POCs, making outcome-based pricing essential for scaling AI adoption.

On what to avoid, he said partners must not get distracted by the rapidly expanding number of AI tools in the market and should instead go deep on a limited set of technologies.

Lazarus also stressed that operating in isolation will no longer work in an AI-driven ecosystem.

“Do not operate in isolation. Partnerships are essential to build complete solutions,” he said.