AI is already making a tangible difference in operations. A significant 60% of operations leaders are currently leveraging AI for intelligent document processing, freeing up valuable time and resources. But this is just the beginning. The vision extends far beyond, with plans to expand AI’s reach into crucial areas like workflow analysis, fraud detection, and streamlining risk and compliance processes. Imagine AI optimising transportation routes in real-time, predicting equipment maintenance needs before they arise, or automating complex scheduling tasks. This is the operational reality AI is creating.
Real-World Impact, Real-World Examples
The impact of AI is not just theoretical. Operations leaders are witnessing firsthand how AI is driving tangible improvements. “With AI-powered vision and sensors, we’ve boosted efficiency, accuracy, and safety in our manufacturing processes,” shares one leader. Others highlight the security benefits: “From fraud detection to claims processing, AI is safeguarding our transactions and improving trust in our services.” Even complex logistical challenges are being conquered: “Our AI-driven logistics solution has cut costs, saved time, and turned complex operations into seamless processes.” These real-world examples showcase the power of AI to deliver concrete results across diverse operational functions.
Operations Takes a Seat at the AI Strategy Table (But Faces Challenges)
With 54% of organisations prioritising cost savings from AI, operations leaders are rightfully taking a seat at the AI strategy table, shaping use cases and driving adoption. A remarkable 56% of operations leaders are actively involved in defining high-value AI applications. However, a disconnect exists. Despite their influence on AI strategy, only a small fraction (7%) of operations leaders have direct data governance responsibilities. This lack of control over the very fuel that powers AI – data – creates a significant hurdle.
Further challenges include data access across siloed systems, limiting the ability to gain a holistic view, difficulty in identifying and prioritising the most impactful AI use cases, and persistent skills shortages. These barriers, while significant, are not deterring operations leaders.
The Future is AI-Driven
Despite these challenges, operations leaders are doubling down on AI. A striking 7 out of 10 plan to prioritise AI investments in 2025, driven by the pursuit of greater cost savings. And the biggest data effort on the horizon? Identifying and prioritising better use cases for AI. This focus on practical applications demonstrates a clear understanding: the future of operations is inextricably linked to the power of AI. By addressing the challenges they face and focusing on strategic implementation, operations leaders are poised to unlock the full potential of AI and transform their organisations.
AI is no longer a buzzword, but Asia Pacific’s transformation engine. It’s reshaping industries and fuelling growth. Initially, high costs and complex ROI pushed leaders toward quick wins. Now, the game has changed. As AI adoption matures, the focus is shifting from short-term gains to long-term, innovation-driven strategies.
GenAI is is at the heart of this shift, moving beyond the periphery to power core business functions and deliver competitive advantage.
Organisations are rethinking AI investments, looking beyond pure financials to consider the impact on jobs, governance, and data readiness. The AI journey is about balancing ambition with practicality.
2. Optimising AI: Tailored Open-Source Models
Smaller, open-source, and specialised AI models will gain momentum as organisations seek efficiency, flexibility, and sustainability in their AI strategies.
Unlike LLMs, which require high computational power, smaller, task-specific models offer comparable performance while being more resource-efficient. This makes them ideal for organisations working with proprietary data or limited computational resources.
Beyond cost and performance, these models are more energy-efficient, addressing growing concerns about AI’s environmental impact.
3. Centralised Tools for Responsible Innovation
Navigating the increasingly complex AI landscape demands unified management and governance. Organisations will prioritise centralised frameworks to tame the chaos of diverse AI solutions, ensuring compliance (think EU AI Act) while boosting transparency and security.
Automated AI lifecycle management tools will streamline oversight, providing real-time tracking of model performance, usage, and issues like drift.
By using flexible developer toolkits and vendor-agnostic strategies, organisations can accelerate innovation while maintaining adaptability, as the technology evolves.
4. Supercharging Workflows With Agentic AI
Organisations will embrace Agentic AI to automate complex workflows and drive business value. Traditional automation tools struggle with real-world dynamism, but AI-powered agents offer a flexible solution. They empower autonomous task execution, intelligent decision-making, and adaptability to changing circumstances.
These agents, often using GenAI, understand complex instructions and learn from experience. They collaborate with humans, boosting efficiency, and adapt to disruptions, unlike rigid traditional automation.
Agentic workflows are key to redefining work, enabling agility and innovation.
5. From Productivity to People
The focus of AI conversations will shift from simply boosting productivity to using AI for human-centric innovation that transforms both employee roles and customer experiences.
For employees, AI will handle routine tasks, enabling them to focus on creativity and innovation. Education and training will be crucial for a smooth transition to AI-powered workflows.
For customers, AI is evolving to offer more empathetic, personalised interactions by understanding individual emotions, motivations, and preferences. Organisations are recognising the need for transparent, explainable AI to build trust, tailor solutions, and deepen engagement.
Hit or miss AI experiments have leaders demanding results. In this breakneck AI landscape, strategy and realism are your survival tools. A pragmatic approach? High-impact, achievable goals. Know your capabilities, prioritise manageable projects, and stay flexible. The AI winners will be those who champion human-AI collaboration, bake in ethics, and never stop researching.
Our research reveals a fascinating dynamic in HR. While 54% of HR leaders currently use AI for recruitment (scanning resumes, etc.), their vision extends far beyond. A striking majority plan to expand AI’s reach into crucial areas: 74% for workforce planning, 68% for talent development and training, and 62% for streamlining employee onboarding.
The impact is tangible, with organisations already seeing significant benefits. GenAI has streamlined presentation creation for bank employees, allowing them to focus on content rather than formatting and improving efficiency. Integrating GenAI into knowledge bases has simplified access to internal information, making it quicker and easier for employees to find answers. AI-driven recruitment screening is accelerating hiring in the insurance sector by analysing resumes and applications to identify top candidates efficiently. Meanwhile, AI-powered workforce management systems are transforming field worker management by optimising job assignments, enabling real-time tracking, and ensuring quick responses to changes.
The Roadblocks and the Opportunity
Despite this promising outlook, HR leaders face significant hurdles. Limited exploration of use cases, the absence of a unified organisational AI strategy, and ethical concerns are among the key barriers to wider AI deployments.
Perhaps most concerning is the limited role HR plays in shaping AI strategy. While 57% of tech and business leaders cite increased productivity as the main driver for AI investments, HR’s influence is surprisingly weak. Only 20% of HR leaders define AI use cases, manage implementation, or are involved in governance and ownership. A mere 8% primarily manage AI solutions.
This disconnect represents a massive opportunity.
2025 and Beyond: A Call to Action for HR
Despite these challenges, our research indicates HR leaders are prioritising AI for 2025. Increased productivity is the top expected outcome, while three in ten will focus on identifying better HR use cases as part of a broader data-centric approach.
The message is clear: HR needs to step up and claim its seat at the AI table. By proactively defining use cases, championing ethical considerations, and collaborating closely with tech teams, HR can transform itself into a strategic driver of AI adoption, unlocking the full potential of this transformative technology for the entire organisation. The future of HR is intelligent, and it’s time for HR leaders to embrace it.
Traditional Conversational AI has faced persistent challenges that have hindered its widespread adoption. Many solutions lack contextual awareness, limiting their ability to engage proactively. Siloed back-end data often restricts these systems from making autonomous decisions, while predefined conversational boundaries prevent seamless, natural interactions.
Despite advancements, organisations deploying Conversational AI continue to encounter significant issues:
Customers frequently need to rephrase or repeat themselves due to misunderstood intent.
Incorrect options frustrate users, pushing them to call contact centres.
Many interactions only partially resolve issues, leaving 40-50% of problems unsolved.
These limitations have slowed adoption, particularly in the Asia Pacific region, where enterprises remain cautious, opting for pilots and tests over large-scale deployments.
Adding to the complexity is the challenge of handling local languages like Thai, Bahasa, Chinese, and Indian languages, as well as nuanced regional English dialects, which AI often struggles to interpret accurately.
Agentic AI: A Transformational Solution
Agentic AI is poised to revolutionise Conversational AI by addressing these longstanding challenges. Unlike traditional systems, Agentic AI offers the ability to retrieve precise information, engage in intelligent, human-like conversations, and make autonomous decisions based on vast amounts of customer metadata.
Agentic AI empowers enterprises to create conversational flows that are not only seamless but also adaptive to context and behaviour.
It enables CX systems to overcome language barriers, handle unstructured data dynamically, and deliver faster, more personalised responses. By doing so, Agentic AI enhances customer satisfaction, drives efficiency, and unlocks the potential for proactive, intelligent engagement at scale.
Success Stories and Adoption Trends
Simpler use cases like balance checks, order confirmations, and structured dialogues have garnered positive feedback. Improvements have been achieved through better conversational design and integrating diverse data into unified repositories. Agent Assist solutions have seen strong adoption in 2024. New developments in AI agents as a digital workforce are unlocking remarkable outcomes. These agents can analyse unstructured CX data, enabling faster, context-rich conversations. In 2025, AI agents with agentic capabilities will make independent decisions, learn from context, solve complex problems, and adapt dynamically based on customer interactions.
Preparing For What’s Ahead
CX solution buyers and decision-makers must prepare for the transformative potential of Agentic AI.
Evaluate vendor offerings. Ask vendors about their Agentic AI solutions and assess their capabilities in delivering desired outcomes.
Look for end-to-end platforms. Ensure platforms provide tools to design, build, test, deploy, and scale AI agents, workflows, and GenAI applications.
Focus on orchestration. Choose solutions that integrate seamlessly across channels and applications, ensuring alignment with voice and human collaboration tools.
The suspected sabotage of 11 undersea cables in 15 months has alarmed NATO countries, prompting increased surveillance around Europe. Patrols will focus on protecting critical assets like fibre optic cables, power lines, gas pipelines, and environmental sensors. Dubbed Baltic Sentry, the mission will deploy frigates, patrol aircraft, and unmanned naval drones, supported by NATO’s Maritime Centre for the Security of Critical Undersea Infrastructure. An AI system will monitor unusual shipping activity, such as loitering near cables or erratic course changes, aiming to cut response times to 30-60 minutes. Meanwhile, Operation Nordic Warden will analyse satellite imagery, patrol data, and Automatic Identification System (AIS) signals to assess risks in 22 key areas.
The primary concern is damage to infrastructure in the shallow waters of the Baltic Sea, but suspicious activity elsewhere has caught the attention of tech giants. Ireland, a critical hub for Europe’s cloud data centres, hosts undersea cables owned by companies like Google, Microsoft, and Amazon, linking it to the US and UK. As a non-NATO country, Ireland faces the challenge of monitoring over 3,000km of coastline. Recently, both the Irish Defence Forces and Royal Navy shadowed a Russian spy ship in the Irish Sea and English Channel. While cable damage is often immediately evident, the risk of communication taps is more alarming and harder to detect.
How Resilient Are Undersea Cable Networks?
There are about 400 undersea cables spanning over 1.3 million kms globally. According to the International Cable Protection Committee, around 200 incidents of cable damage occur annually, mostly caused by dragged anchors or trawling. Only about 10% result from natural causes like weather or wildlife. Near shorelines, cables are heavily protected and often buried under several metres of sand in shallow waters. However, in deeper seas, they are harder to monitor and safeguard.
Highly developed regions, such as the Baltic Sea, North Sea, and Irish Sea, rely on multiple redundant cables to maintain connections between countries. While severing a single link may reduce capacity and cause inconvenience, major disruptions are rare, even for remote European islands served by multiple cables.
Fibre optic cable repairs typically take days to weeks, faster than the lengthy timelines for fixing power cables or gas pipelines. Repair costs range from USD 1-3 million depending on the damage. Faults are located using test pulses, and specialised ships lift the damaged sections to the surface for splicing. However, with only 22 repair-designated cable ships worldwide, simultaneous outages could significantly delay restoration.
In regions with less cooperative neighbours, obtaining permissions can further slow repairs. For instance, cables crossing the South China Sea face increasing challenges in deployment and maintenance, complicating connections between ASEAN nations. Routing cables along longer coastal paths raises costs and impacts latency, adding further strain to the network.
Responding to Escalating Incidents
Plausible deniability and the opaque nature of maritime operations make attributing these events challenging. Nonetheless, NATO countries view them as part of Russia’s broader hybrid warfare strategy, which avoids direct confrontation while instilling fear and uncertainty by showcasing an adversary’s reach. Attacks on undersea cables undermine public trust in a government’s ability to protect critical infrastructure.
European governments initially downplayed the impact of these attacks, likely to minimise psychological effects and avoid escalation. While this cautious approach, coupled with rapid repairs, proved effective in the short term, it may have emboldened adversaries, leading to further incidents. In response, Sweden and Finland are now more willing to seize vessels in their territorial waters to deter both intentional and negligent actions.
Implications for Enterprise Networks
While enterprises cannot prevent damage to undersea infrastructure, they can mitigate risks and build resilient networks:
Satellite Connectivity. Satellite internet services like Starlink and Eutelsat may not be ideal for bandwidth-intensive applications but can support critical services requiring international connections. An SD-WAN enables automatic failover to a redundant circuit if a land-based or undersea cable is disrupted.
Dynamic Path Selection. Modern WAN architectures with dynamic path selection can reroute traffic to alternate cloud regions when primary paths are down. Locally available services can continue operating on domestic networks unaffected by international outages.
Edge Computing. Adopting an edge-to-cloud strategy allows the running of select workloads closer to the edge or in local data centres. This reduces reliance on international links, improves resilience, and lowers latency.
Disaster Recovery Planning. Enterprises should incorporate extended network outages into their disaster recovery plans, assessing the potential impact on operations and distinguishing between land-based, undersea, and other types of connections.
The recent unveiling of Project Stargate sent ripples throughout the tech world, not just for its ambitious goals, but for its staggering price tag: a cool USD 500B over four years. Let that sink in. That’s roughly the equivalent of Singapore’s entire GDP in 2023. For context, that kind of money could fund the entire Apollo programand build twoInternational Space Stations, with some spending money left over. It’s a figure that underscores the sheer scale of investment required to push the boundaries of AI.
But then, the plot thickened. A relatively unknown Chinese company, DeepSeek, seemingly out of nowhere, launched its R1 large language model (LLM). Not only does R1 appear to be a direct competitor to OpenAI’s latest offerings, but DeepSeek also claims to have achieved this feat at a fraction of the cost, and using fewer (and potentially less powerful) GPUs. This announcement sent shockwaves through the stock market on January 27th, impacting nearly every stock associated with AI chip manufacturing. Nvidia (NVDA), a key player in the AI hardware space, suffered one of the biggest single-day losses in US stock market history, with nearly USD 600B wiped off its market capitalisation. Ironically, that’s more than Project Stargate’s entire budget plus the cost of an ISS.
This dramatic market reaction highlights several critical trends emerging in 2025. The previously observed consensus on AI risks and legislation is already beginning to fracture (witness the recent back-and-forth on AI regulation). Meanwhile, the exorbitant cost of AI development is becoming increasingly apparent. We’re also seeing a renewed West versus (Far) East rivalry playing out in the AI arena, extending beyond just technological competition. And finally, the age-old debate between open-source and proprietary software is back, with some LLMs, like DeepSeek’s R1, leaning more towards open access than others.
For organisations considering investing in AI, and indeed for all of us whose lives are increasingly touched by AI developments, it’s crucial to keep a close watch on these powerful trends. The risks, the investments, and the potential benefits of AI must be carefully scrutinised and potentially reassessed. The recent stock market correction suggests a necessary pushback against the over-confidence and over-spending that has characterised some areas of AI development. As DeepSeek’s R1 has shown, sometimes it doesn’t take much to disrupt the party.
The question now is: how will the landscape shift, and who will emerge as the true leaders in this expensive, yet potentially transformative, race?