From potential to impact: Harnessing AI in 2025

AS 2025 unfolds, mid-market organisations are poised to lead the artificial intelligence (AI) revolution in Asia-Pacific. These businesses form a dynamic and vital segment of the region’s economy, being agile, ambitious and adept at leveraging new technologies to fuel growth. However, their reinvention – to accelerate growth and build resilience – will depend on their ability to seize defining moments and key opportunities in the coming year.

Trend one: Redefining measurement of AI ROI

The way organisations measure AI’s value is undergoing a transformation.

While many mid-market organisations are increasing their AI budgets and anticipating up to a fourfold return on investment within a year, the reality is more nuanced. Traditional metrics such as headcount reduction – used by just 9 per cent of global decision-makers – fail to capture the broader impact of AI.

In 2025, leaders will need a more holistic approach to evaluating the value of AI. This means moving beyond simplistic measures and embracing new metrics that reflect the changing nature of work.

AI is transforming not only how tasks are completed, but also redefining the essence of work itself. Organisations that align AI initiatives with meaningful business outcomes – not just cost efficiency – will be better positioned to sustain productivity gains and unlock AI’s full potential.

Trend two: Balancing short-term gains with long-term vision

The urgency to adopt AI is undeniable – 85 per cent of mid-market leaders in Asia-Pacific fear losing their competitive edge without rapid adoption.

This pressure has led many to fast-track implementations, often through fragmented, “out-of-the-box” approaches that lack strategic unity. Consequently, such hasty deployments can undermine AI’s impact and hinder scalability.

To succeed, leaders must balance speed with a long-term perspective. A cohesive AI road map will act as a “north star”, ensuring adoption is both scalable and sustainable. When developing a holistic AI plan, technology can act as a good starting point, but organisations need to recognise that people and data are equally critical.

AI fluency must extend across the organisation, with upskilling and training initiatives that empower employees to fully embrace the technology. A robust data strategy is also essential.

SEE ALSO

AI alone won’t save the planet

The new frontier of Asia’s venture capital landscape in the age of AI and geopolitical shifts 

Harmonising data from diverse systems and formats is critical for AI to deliver accurate and reliable insights. Trust in AI hinges on the quality of the data it uses, and organisations without confidence in their data risk undermining all their AI initiatives.

Trend three: Building trust in AI

Trust will be the cornerstone of AI’s success in 2025. While 96 per cent of mid-market organisations globally are using AI to make decisions, trust in AI outputs has decreased – from 48 per cent in 2023 to 26 per cent in 2024. This decline stems largely from concerns about data quality and the reliability of AI-driven insights.

Rebuilding this trust involves setting clear guard rails for AI usage, determining where human oversight is necessary, and addressing data quality issues through robust governance frameworks.

An important step in responsible AI involves teaching systems to “forget” – ensuring models can unlearn biases or incorrect information. Encouragingly, 96 per cent of mid-market organisations globally plan to fine-tune or retrain their AI models by the end of 2025, to omit or unlearn certain knowledge.

Leaders can foster trust and accelerate AI adoption by creating sandbox environments, where employees can safely experiment with AI on non-critical tasks. Encouraging exploration within the framework of responsible AI guidelines and compliance reinforces the value of calculated risks.

Implementing an employee feedback system further promotes open communication and continuous learning, in turn building trust and shortening the path from innovation to competitive advantage.

Looking ahead

Asia-Pacific’s mid-market organisations stand at a critical juncture. With tremendous potential to become industry titans of tomorrow, these businesses are embracing AI with enthusiasm, confidence and ambition.

To fully realise AI’s value, mid-market leaders must rethink how success is measured, balance short-term gains with a cohesive long-term strategy, and prioritise trust at every step of the journey. By addressing these trends head-on, mid-market organisations will not only compete but lead – shaping the future of Asia-Pacific’s rapidly evolving digital economy.

The writer is president, Asia-Pacific, at Avanade