The relationship between investors and companies has never been simple, but it has always depended on one thing: the quality of the conversation. For decades, that conversation happened in earnings calls, annual meetings, one-on-one roadshows, and the occasional frank exchange between a portfolio manager and a CEO. The substance of those conversations (the tone, the context, the judgment calls made in real time) was what distinguished the investors who really understood a company.

AI is now changing the structure of that engagement. At a recent FCLTGlobal forum convening senior leaders from asset management, investor relations, and corporate development, what emerged was a set of specific practices organized around three shifts:

• Companies are rethinking how they communicate to humans and to AI agents alike.
• Investors are strengthening the AI-based tools they use to test conviction and govern engagement.
• The sell-side is repositioning its value proposition from speed to judgment.

Each shift builds on the one before it: as companies change what and how they disclose, investors change how they interpret and act on it, and the intermediaries between them are forced to redefine what value they provide. Together, they form a single picture of how investor-corporate engagement itself is being rewritten.

Companies are rethinking how they communicate

Disclosure now has two audiences: humans and agents

Corporate communications have always been crafted with analysts, portfolio managers, and sell-side researchers in mind. Today, that audience also includes AI models that will scrape, summarize, and redistribute the content before most human readers have opened it.

Investor relations teams have begun adjusting accordingly. One head of IR at a major investment bank described a deliberate decision to add macroeconomic context to quarterly earnings scripts because an AI model summarizing the call six quarters later would lack it entirely. Without that context, a future analysis of how the company had performed across different market environments would be missing the interpretive frame that made the original disclosure meaningful.

The practice extends further, including into regulatory disclosure. Some companies are beginning to use AI tools to draft and review the narrative sections of their filings, including Reg S-K disclosures, testing whether the language holds up not just to a human reader but to an AI model asked to summarize the company’s risk profile or strategic position from the filing alone.

IR professionals are also running their scripts through AI tools before publication, asking the model to summarize the key takeaways for any investor who did not attend the call. But verifying that the model is producing what the company actually intends to communicate is an essential human step.

Proprietary data as a differentiator

As AI models become standard tools for investment analysis, the companies that stand out are not necessarily those with the best public disclosures. They are the ones with information that no one else has. In a world where every investor is running the same models on the same public information, proprietary data is one of the few sources of differentiation.

One global holding company described a deliberate effort to capture institutional knowledge that had historically lived in individuals rather than systems: the judgment calls, relationship histories, and contextual insights that experienced executives carry with them but rarely formalize. A legal team working on a complex acquisition can draw on a decade of related transactions. A business development team evaluating a new market can access pattern recognition that would otherwise take years to accumulate.

For investors, the implication is significant. As this kind of proprietary data infrastructure becomes more common among leading companies, the ability to access and interpret it through direct engagement, rather than public filings alone, is becoming an increasingly important advantage.

Agility is becoming a governance principle

Perhaps the most important shift visible among leading companies is a change in how they think about the pace of adaptation itself. In an environment where AI capabilities are advancing faster than most institutions can adapt their policies, the companies navigating this most effectively are those that have built agility into their governance structures rather than treating it as a cultural aspiration.

This means tracking what AI tools are doing inside the organization: on the technology infrastructure side, on cost, and on decision quality, and being willing to adjust quickly when the evidence warrants it. A framework needs to be in place for evaluating new tools before the pressure to adopt them becomes acute. One company at the forum described doing exactly this: before deploying any new AI tool, they defined upfront what success looked like on both the cost and decision-quality dimensions, with explicit checkpoints to reassess.

Investors are refining how they use AI

AI is being used to stress-test the bear case

One of the oldest disciplines in investment management is the practice of seeking out the strongest possible argument against a position you hold. AI has given investors a practical tool for doing it more systematically. Rather than asking a model for a summary of a company’s financials, sophisticated investors are now asking it to adopt a specific point of view: to analyze a holding the way an activist would, or to build the “bear case” the way a short seller might. Some are going further, asking models to take on “named investor personas,” assessing a position the way a value investor known for a particular methodology might, or identifying the vulnerabilities a distressed investor would target.

The value of this practice lies in the discipline it creates. Investors who build this kind of adversarial analysis into their process are less likely to be surprised by the arguments that eventually surface in the market and are better positioned to engage with companies on the strategic questions that determine long-term value.

Governance teams are breaking down internal silos

Governance professionals and portfolio managers often operate in separate lanes, making decisions about engagement and voting without a shared framework, creating visible inconsistencies in how they communicate with companies and how they behave at shareholder meetings. As AI-driven analysis becomes more capable of synthesizing years of voting history, public statements, and engagement records into a single pattern, those inconsistencies become far more apparent to outside observers than they once were.

Several participants at the forum described active efforts to resolve these silos: joint frameworks for escalation, shared decision-making processes for override situations, and clearer lines of accountability between stewardship teams and investment professionals.

The sell-side is repositioning its value proposition

Judgment is replacing speed as the core offering

For much of the past two decades, the sell-side’s primary value proposition rested on the ability to process public information quickly and distribute a summary to clients before competitors did. That value proposition is now under significant pressure since AI tools can summarize an earnings call, compile analyst estimates, and identify consensus revisions faster than any research team, at a fraction of the cost.

What AI cannot yet replicate is judgment about what matters and why. The sell-side analysts navigating this transition successfully are shifting their focus from information aggregation toward interpretation: scenario analysis, thematic research, bespoke deep dives on questions that institutional clients are actually wrestling with. One asset manager at the forum described a technology company that had deployed AI to screen 20 sell-side reports on a single name each morning, surfacing only three that offered a genuinely differentiated view. The reports that made it through that filter were not the fastest, but rather the ones with a point of view worth paying for.

The conversation that AI cannot replace

None of the practices described above requires predicting where AI will be in five years. Each one is available now, and each one reflects a version of the same underlying insight: the human conversation between investors and companies is not being replaced by AI, but it is being restructured around it.

The companies and investors who are adapting well are not treating AI as a new condition of the engagement environment, one that raises the stakes for the quality of direct dialogue, and that rewards the institutions that have invested in the relationships, the frameworks, and the organizational discipline to make that dialogue count.