Consider a scenario where a wealth manager is preparing a significant co-investment opportunity for a longstanding client. The client is well-capitalised and has a clean, professional history. The counterparty’s due-diligence team runs a quick search on ChatGPT.
The AI claims the client had a history of legal trouble – the claim tracing back to an old Reddit thread and a misinterpreted Wikipedia entry. No further verification takes place. And the counterparty quietly withdraws from the table without offering a specific reason.
Situations like this illustrate a new and increasingly relevant risk for investment professionals: AI-generated reputation. Advisers may traditionally focus on market volatility and complex regulatory compliance requirements but AI-generated reputation now demands similar attention.
This shift represents a fundamental change in how the world perceives wealthy individuals. Reputation is no longer just a collection of press links and search results – in fact, it is now a synthesised story created by large language models without editorial judgement or contextual verification.
The move from search to stories
For years Google has been the main portal for credible background research. Users would scroll through various links and evaluate the source’s credibility individually. For many professional investment teams, however, that evaluative process is now giving way to something faster and less transparent.
AI search platforms now combine information into a single, authoritative-sounding narrative for the user. ChatGPT currently receives approximately six billion monthly visits as users seek direct answers. Google Gemini also demonstrates massive scale, with 750 million monthly active users globally.
“Wealth managers can add genuine value by integrating reputation audits into their regular reviews.
Wealthy individuals who once relied on privacy through obscurity are now far more visible in AI-generated searches.”
This transition from traditional search to AI-led synthesis changes the nature of due-diligence. Technology consultant Gartner predicts organic search traffic will decline by 50% or more by 2028. The shift also carries economic implications, with Semrush research predicting AI-powered search could match or exceed the economic value of traditional Google search by 2027.
Wealthy individuals who once relied on privacy through obscurity are now far more visible in AI-generated searches. AI systems do not recognise the privacy preferences of many wealthy families. If a client lacks a strong digital presence, the AI will fill the void. It will pull from any available corner of the internet to construct its response.
The data provenance problem
Wealth managers often assume AI tools rely primarily on verified news sources or official records. In reality, the reliability of any AI-generated summary depends entirely on the data the model has been trained on.
Research from Yext indicates some 86% of AI citations come from controllable sources – however, this figure is only relevant if those controllable sources actually exist online. Many high-net-worth individuals purposefully keep their official biographies and company details very lean. This lack of authoritative data in turn forces AI models to look toward uncontrolled digital spaces.
Profound’s analysis of 30 million citations shows that Wikipedia remains a dominant primary source – indeed, it accounts for nearly half of the most-cited sources within the ChatGPT ecosystem. Semrush research meanwhile indicates Reddit appears in approximately 40% of all model responses. Remember – both platforms are often edited by anonymous users or contain unverified personal opinions.
Ultimately, notes Semrush, approximately 90% of ChatGPT citations originate from pages ranking 21st or lower in search. In other words, AI models find the information that traditional search engines would never have placed in front of a researcher.
‘Eight-to-one’ imbalance
Research conducted by Pavesen found that uncontrolled sources outnumber controlled sources in online data environments by approximately eight-to-one for many private individuals. Controlled sources include official firm websites, authorised biographies, and verified professional social media profiles.
Uncontrolled sources, meanwhile, range from news archives and blogposts to forum discussions and historical legal records. For AI systems, this imbalance matters. The model has significantly more third-party material to draw on than first-party information from the client.
The result is predictable. AI narratives tend to emphasise the most abundant material rather than the most accurate or relevant. A single negative news article from 20 years ago can become a central theme. The narrative becomes a permanent fixture of the client’s digital profile across all platforms. Advisers must treat this imbalance as a threat to their client’s future deal flow.
If an AI suggests a risk, a potential partner may simply move on. They rarely take the time to verify if the AI ‘hallucinated’ the details.”
Reputation risk has direct implications for co-investments, counterparty trust and even regulatory perception. Investment professionals and fund selectors increasingly use AI to perform preliminary background checks.
If an AI suggests a risk, a potential partner may simply move on. They rarely take the time to verify if the AI ‘hallucinated’ the details. The result is a silent filter on deal flow. One that neither the client nor their adviser may ever detect without proactive monitoring.
There is also a significant overlap between AI reputation risk and broader cybersecurity concerns. Deloitte reports that 43% of family offices globally have experienced a cyberattack in recent years. For those managing more than $1bn (£750m), the figure rises significantly to 62% of offices.
Omega Systems found that 83% of family offices are worried about sophisticated deepfake campaigns. A successful data breach can provide fresh, negative material for AI models to cite. This misinformation can then become a permanent part of the client’s synthesised digital identity.
Proactive management strategies for advisers
Wealth managers can add genuine value by integrating reputation audits into their regular reviews. The first step involves querying platforms such as ChatGPT, Gemini, Perplexity, Claude and Microsoft Copilot. Advisers should analyse the specific information these models provide about their key clients. They must check which sources are being cited and identify any factual inaccuracies.
This process helps the adviser understand how a counterparty might initially perceive the client. Monitoring these outputs is now just as important as monitoring a credit report. Advisers should also encourage clients to develop more authoritative first-party content online.
This does not mean launching a public relations campaign for the sake of vanity but, rather, ensuring accurate, verified data is available for AI systems to crawl. Placing thought leadership in reputable publications creates high-quality citations AI models will trust. Increasing the volume of controlled data helps rebalance the eight-to-one ratio. Proactive placement of accurate information is the best defence against AI-generated reputation risk.
The range of risks wealth managers must consider has broadened significantly in recent years. Alongside market, credit and regulatory exposure, AI-generated reputation risk is emerging as a practical factor in deal-making and client perception. It can impact liquidity by blocking exits or preventing new partnerships from forming. It can also complicate the onboarding process with new banks or legal firms.
Ignoring the digital narrative is no longer an option for serious investment professionals. Understanding how AI systems gather and interpret information allows wealth managers to spot reputational risks before they quietly affect investment relationships or deal flow. As AI increasingly shapes first impressions in professional due-diligence, digital reputation has become another factor advisers must actively manage.
Tony McChrystal is the founder of Pavesen, a London-based reputation management firm advising high-profile individuals, family offices and C-suite executives on reputation risk and digital footprint strategy.
