Why Singapore companies need to manage AI-mediated reputation

GENERATIVE artificial intelligence (GenAI) is changing more than how information is created; it is changing how companies are understood.

For years, reputation management followed a familiar model: companies shaped messages, engaged stakeholders, monitored coverage and adjusted when necessary.

Stakeholders largely formed opinions through a combination of direct and indirect channels such as news reports, company statements, key influencers, analyst commentary and lived experience.

That dynamic is evolving rapidly with GenAI.

Stakeholders are increasingly turning to large language models to understand companies, industries and issues. AI is no longer just accelerating the flow of information.

It is interpreting it – synthesising sources, compressing narratives and shaping how organisations are perceived in real time.

In effect, AI is becoming a new intermediary in corporate reputation.

This shift matters everywhere, but particularly in Singapore, where trust carries strategic weight. Companies are judged on visibility, but also on governance, credibility and consistency across investors, regulators, employees, customers and regional partners.

When stakeholders ask AI systems about a company, they are now given a response that moves beyond simply retrieving facts. They are receiving machine-generated interpretations built from multiple sources and assumptions.

Those summaries can be useful, but they can also be incomplete, outdated or skewed towards the signals an AI model prioritises.

That creates a new category of reputational risk and communications complexity.

Recent studies have shown that over 50 per cent of consumers already use large language models to gather product information, while 60 per cent do not click through to original sources afterwards.

Other critical stakeholders are also increasingly relying on AI-generated summaries rather than engaging directly with company material.

For businesses, the implication is clear: it is no longer enough to ensure a message exists.

Companies now need to understand how AI systems are interpreting that message, which signals are being amplified and whether the resulting narrative reflects the organisation they are trying to build.

This requires a shift in mindset across three areas.

First, companies need visibility into how AI currently represents them. That starts with asking the same questions stakeholders would ask, and reviewing how different models describe the organisation, its leadership and its competitors.

Such exercises often reveal inconsistencies between intended positioning and AI-generated perception.

Second, companies need to focus less on volume and more on narrative relevance. AI systems flatten and compress information, which means nuance can easily disappear.

In Singapore, these differences matter. Issues such as sustainability, workforce decisions, geopolitical exposure and corporate conduct are interpreted differently by investors, policymakers, employees and the public.

A generic AI summary may blur those distinctions. Companies that fail to understand where stakeholder concerns diverge risk responding too slowly – or responding to the wrong issue altogether.

This is where more targeted monitoring becomes essential. Organisations need to identify which narratives are shaping stakeholder perceptions, which sources AI systems rely on most heavily and where reputational risks are beginning to gain traction.

Third, organisations need to move from episodic reputation management to continuous adaptation. Reputation can now shift quickly as AI systems absorb new information and recirculate it across audiences at scale.

That means reputation management can no longer operate as a periodic communications exercise. It requires ongoing monitoring, rapid interpretation and cross-functional coordination between communications, public affairs, investor relations and leadership teams.

The challenge is increasingly strategic, not simply operational.

Business leaders today need to be able to anticipate how narratives may evolve before they harden into perception informed by GenAI.

For companies operating from Singapore, AI-mediated reputation should sit alongside enterprise risk, investor confidence and stakeholder engagement as a board-level concern.

Three questions deserve immediate attention: How does AI currently describe our organisation? Which narratives are gaining traction across key stakeholder groups? And who inside the organisation is responsible for monitoring and responding as those narratives evolve?

The broader shift is straightforward. Reputation is no longer formed only through what companies say and what stakeholders read. Increasingly, it is shaped by what machines infer.

Companies that adapt early will be better positioned to influence how they are understood. Those that do not risk having their reputation defined by systems they neither control nor fully see.

Both writers are from The Penta Group. Dan La Russo is senior partner and Shawn Balakrishnan is partner