On 12th November 2025, The Barton Partnership hosted a Private Equity roundtable in Singapore with senior Private Equity investors, value-creation leaders, and portfolio company executives from across Asia-Pacific. The discussion focused on how AI is impacting Private Equity firms and portfolio companies, shifting from AI pilots to practical use cases that directly impact operational performance and financial outcomes.
Download as a PDF
Participants shared perspectives on the operational, strategic, and human challenges of embedding AI at scale. What emerged was a pragmatic view: AI enthusiasm is high, but value creation now hinges on disciplined prioritisation, credible measurable outcomes, and sustained capability building. From a Private Equity investing perspective, AI is reshaping deal evaluation.
AI is a focal discussion topic amongst every PE Investment Committee, as it will impact existing business models. It is important to evaluate which business models can be enhanced rather than disrupted by AI.
#1 From Pilots to P&L
AI has moved beyond experimentation to measurable application, though many organisations are still learning where and how to scale.
Rather than running numerous disconnected pilots, firms are concentrating resources on a small number of high-impact use cases. Examples include pricing optimisation, forecasting and planning, alongside efficiency improvements in data-heavy functions.
While productivity tools such as coding or document assistance are widespread, they rarely shift the P&L alone. The challenge lies in linking AI initiatives directly to revenue or margin improvement. Participants agreed that clear KPIs, baselines and ownership are essential to attributing results to the EBIDTA or revenue upside.
The discussion emphasised the importance of rigorous measurement, with value attributed explicitly tooutcomes that
#2 AI Adoption and Enablement
Momentum is growing from both directions: bottom-up experimentation by teams and top-down expectation from boards and investment committees.
Spreading efforts and resources across one-off use cases throughout the company or focusing on immediate ROI may result in lack of sustained progress, and organisational fatigue. Ultimately, there are a 3-5 use cases that matter most in any given company that move the needle, which highlights the need for companies to spend more time prioritising the right areas before jumping to execution.
Finance, HR, and corporate functions have often been the biggest use cases for AI, while investment teams are moving more cautiously, given the cost AI deployment, and potential analytical errors. Effective adoption depends on three enablers:
Training and change management: Structured enablement, consistent communication, and peer learning are essential. Uncoordinated tool rollouts often stall without behavioural reinforcement. Tying KPIs and compensation to AI adoption also aligns incentives
Poor data quality remains the greatest barrier to scaling. Participants stressed the need for consistent data structures and clear governance so that AI outputs can be repeated and trusted.
Governance and flexibility: Clear guardrails on acceptable use, data handling, and model selection protect organisations while allowing for ongoing experimentation.
Successful firms are institutionalising verification by design, using AI for speed and pattern recognition, but validating results through human judgement, oversight and source checking.
On the infrastructure level, on the supply side of AI, the constraint is not demand, but rather the constraint is chip supply and power. Every GPU that gets deployed goes live immediately.
Whenever a company deploys AI deeply, it is important to protect or enhance an existing competitive strength or to create a new advantage that is hard to copy, while recognising that competition can catch up faster.
#4 Asia Sector Readiness and Regional Realities
Adoption patterns vary by sector and geography. Consumer-facing and service businesses are advancing fastest due to the availability of ready-to-deploy solutions and clearer impacts on revenue outcomes. In contrast, industrial and B2B organisations face slower progress, hindered by legacy systems, manual processes, and limited off-the-shelf tools.
Across Asia-Pacific, participants described differing starting points and operating realities. In several Southeast Asian portfolio companies, AI has been placed on the agenda from the board level, while day-today usage continues to grow bottom-up and can be unstructured. In smaller, more contained environments, greenfield builds can be easier to wire correctly from the outset, which influences the build versus buy calculus and sequencing.
Asian markets force AI adoption to be pragmatic: High costs of AI adoption across fragmented Asian market due to regulatory and language differences as compared to lower costs of labour, and AI talent gaps are the key considerations. What matters more than technology to unlock value is data maturity and access, localisation of AI models and workflows, and adoption in the frontlines.
Participants noted that even if AI does not generate material cost reductions for some companies, but by increasing productivity of the workforce, has resulted in a reduction or capping of cost increase.
These contextual differences led participants to emphasise tailoring scope and sequencing to local conditions and sector maturity, rather than copying standardised approaches from elsewhere.
#5 Talent, Culture, and the Changing Workforce
AI is reshaping both hiring priorities and workplace expectations. Curiosity, adaptability, and AI literacy are becoming core leadership attributes, as organisations seek individuals who can interrogate, verify, and apply AI-generated insights responsibly.
Traditional PE playbook shifting increasingly from governance and efficiency to AI powered growth: AI driven revenue acceleration (lead gen, pipe maturity, churn, pricing), AI enabled propositions taking centre stage in Asia vs. efficiency / S&A play which are table stakes. PE firms using AI to accelerate deal cycles, Proactive investments from IC on AI talent and investments in AI led growth.
Traditional consulting pyramids may evolve toward more diamond-shaped structures, as automation removes routine junior work and experienced managers take on a greater share of solution design and client delivery.
Approaches to training are also shifting, with firms moving away from passive tool demonstrations toward role-based learning and practical scenarios.
Recruitment processes are adapting too. Instead of attempting to eliminate AI from candidate assessments, organisations are testing applicants’ ability to use AI effectively and transparently, introducing dialogue-based case studies and live problem-solving exercises to assess reasoning and verification skills.
#6 Measuring Value, Managing Risk
The most significant results so far have come from productivity gains and workflow automation, with clear benefits in research, reporting, and document-heavy processes.
To sustain progress, AI programmes must transition from automation to transformation, rewiring end-to-end processes and aligning them to business metrics. Full disruption will take longer, as industry dynamics and product cycles differ widely.
Effective governance remains critical. Organisations are formalising frameworks that define acceptable use, model transparency, and ensure data integrity, treating them as living documents that evolve in response to technology and regulatory changes.
AI hallucination is one of the bottlenecks to large scale adoption, which requires checking with the source and having a level of scepticism and not completely outsource the thinking process.
Success ultimately hinges on sequencing: start with low-risk use cases, rigorously measure the results, and expand once attribution and trust are established.
AI is a permanent feature of the private equity value creation lifecycle, and its real impact depends more on disciplined execution than on technology. The most successful firms are those clarifying ownership, investing in foundations (data, measurement, people), and resisting the urge to chase every shiny use case
Across Asia-Pacific, adoption is pragmatic rather than speculative: ambition balanced by operational realism. The goal is not to automate everything, but to use AI deliberately and credibly to enhance decision-making, productivity, and long-term enterprise value.