AI technologies such as large language model–based assistants, enterprise AI tools, and financial data copilots has led to a revolution in information processing. People now rely more on enterprise AI assistants to help them synthesize data into useful insights, as part of their jobs or management of their investments.
Analysts are increasingly likely to use large language models (LLMs) to summarize filings, compare risk factors, extract executive compensation data, or analyze ESG statements. The most common AI models synthesize a range of financial disclosures, such as Form 10-K, Form 10-Q, proxy statements, earnings transcripts, or investor relations website content.
The trick lies in making the content accurate for machines to read and process. If financial information is unclear, inconsistent, or poorly structured, the AI system may misinterpret it and produce incomplete or inaccurate summaries to shareholders.
Fortunately, there is a way to ensure that you avoid common AI pitfalls while optimizing financial reports. This guide will show you how tailoring shareholder materials for AI and LLMs is beneficial for your company and its investors.
What It Means to Optimize Shareholder Materials for LLMs
While it may seem like optimizing shareholder materials for large language models means changing the structure and content in ways that might affect regulatory compliance, that is not necessarily the case. This type of optimization can easily align with government regulations, as well.
Structuring documentation for LLMs requires attention to structure, clarity, semantic consistency, and data formatting. Focusing on optimization for an AI system might look like:
- Outlining clear headings
- Defining terminology
- Using consistent metrics that rely on GAAP-based metrics
- Creating a logical, easy-to-follow organization
The chief premise here is that companies can create typical disclosures for SEC compliance that are easy for investors to summarize using AI tools. The same capability applies to other investor communications, like certain shareholder proposals or other investor communications.
Key Shareholder Materials to Consider
Making shareholder communication easy to synthesize using language models can take some time, so you should know where to start first. This may involve getting into the heads of your investors. What are they most likely to want summarized into easy insights? In many cases, they are looking to condense financial data into readable forms or take long narratives and sum up the main points. As such, you should begin by optimizing these disclosures for AI usage:
- Governance charters and policies
- Annual reports, such as Form 10-K
- Quarterly reports, like Form 10-Q
- Proxy statements, specifically SEC Form DEF 14A
- ESG disclosures or sustainability reports
- Earnings releases
- Investor presentations
- Investor relations website FAQs
As you make progress through each type of disclosure, aim for structural and semantic consistency between all documentation, where applicable. This effort can increase quality of financial reporting, as well as improving the results of AI use.
Core Principles for AI-Ready Shareholder Materials
To make your shareholder materials easy for AI tools to summarize accurately, you should follow these core principles.
1. Structural Clarity
In order for an AI tool to synthesize information accurately, the documents require structural clarity. Information that is poorly organized or illogical is likely to generate AI results that do not make sense or are actively misleading. Instead, you should ensure that each document has the following features:
- Logical information hierarchy, including consistent headers
- Defined section titles
- Clear, descriptive headings
- Standardized formatting across documents
- Clear labeling and cross-references for tables
Anyone reading the disclosure should be able to identify where they are and what is being discussed at a glance, based on the structure.
2. Terminology Consistency
While varying language can be helpful for engagement, it often creates confusion for AI summaries. Consistency is key to making references readable and easier to synthesize. Follow these best practices:
- Clarify acronyms on first use.
- Define metrics clearly, especially if they are non-GAAP.
- Be consistent in naming segments.
- Avoid alternating terminology, such as referring to Adjusted EBITDA as Adj. EBITDA.
- Use standardized language for ESG terminology.
This consistency helps AI summaries to properly organize the information and avoid leaving anything out.
3. Data Integrity & Structured Reporting
The way that you structure and tag financial data is different for human readers than machines. Machines rely on tags to tell them how the information is organized and what it presents, much like a website uses HTML to know how to present the information in a browser. Tagging is a key component of document structure, for items such as financial statements, risk disclosures, and executive compensation. Proper tagging includes:
- Accuracy with Inline XBRL
- Consistency between narratives and financial tables
- Alignment for MD&A commentary and figures
These steps help to preserve the integrity of the data and final reporting.
4. Narrative Precision
The narrative sections of any disclosure should be highly readable, but also easy for machines to scan. Precision helps to clarify language so that machines doing AI summaries can correctly synthesize it into compact insights. Look for ways that the language could be misinterpreted and seek to clarify. For example, instead of using “significantly more than,” make specific statements like “50% more than.”
Aim to structure documentation for faster processing and higher accuracy, especially when outlining past and present considerations. Avoid using overly promotional phrasing. Instead, let the numbers speak for themselves. When you compare items, use measurable language that is easy to follow. If you compare historical performance and projections, clearly distinguish them. Be sure to align risk factors with business strategies to manage them.
AI-Specific Risks in Shareholder Communication
While artificial intelligence can be a practical tool for data synthesis and sentiment analysis, it is not without flaws. Companies have to take care when optimizing for AI in financial reporting, to ensure that the summaries actually represent accurate information. AI may present the following issues:
- Biased summaries caused by vague risk-factor statements
- Misunderstandings about executive compensation
- Perception gaps from inconsistent ESG terminology
- Inaccurate information due to outdated website content that conflicts with current filings
- Misinterpretation of complex tables as a result of formatting errors
- Generation of fake assessments due to unclear data
Inaccurate information can hurt investor confidence. It may increase regulatory scrutiny, as poor formatting may raise other inconsistencies to the surface.
Practical Steps to Make Shareholder Materials AI-Ready
To ensure that your disclosures are ready for AI synthesis, complete the following steps:
- Standardize terminology across filings, including narrative sections and investor communications.
- Improve document hierarchy and headings.
- Confirm that your Inline XBRL tagging is accurate.
- Align language and messaging on investor relations websites with SEC filings.
- Clearly structure risk factors.
- Ensure executive biographies are consistent across filings and investor communications.
- Create ESG metrics that are measurable and comparable.
- Reduce duplicative statements to avoid overcounting.
- Revise contradictory statements for greater clarity and accuracy.
The ideal system will regularly check these items for accuracy and consistency over time.
AI, Disclosure Controls, and Corporate Governance
Although it is easy to think of a large language model as something that does not need monitoring, AI oversight is critical to the process. You will need to integrate AI review as part of your disclosure controls and governance, including:
- Discussion of AI requirements and limitations in disclosure committees
- Integrated AI awareness into internal reporting workflows
- Alignment of AI understanding and priorities between legal, IR, communications, and finance teams
- Plans for protecting financial data privacy
It helps to look at your company’s digital presentation as part of reputational risk management. Monitor how the organization’s information appears in AI-generated summaries to identify inaccuracies or inconsistencies. This work can help you assess your progress and ensure that investors get better results when they perform the same actions.
The Future of Investor Communications in an AI-First World
People are becoming more comfortable and reliant on technology to perform a series of routine tasks that used to take them hours or days, such as drafting an AI-assisted proposal or summary. They expect data to have a clearly defined structure that is machine-readable. Alongside these developments, businesses have to be ready for the impact that AI technology can present for their organizations and investor communications. Specifically, you need to prepare for AI-driven analyst reports, automated peer benchmarking, and increased scrutiny of the consistency of your disclosures.
The good news is that streamlining your disclosures to work with LLMs can help improve your regulatory compliance as well as investor relationships. By getting ahead of the demand, your company can improve overall clarity, reduce risk due to misinterpretation, and strengthen investor trust.
