Can AI track how corporate narratives change?
What if you could instantly see how a company’s story has changed from one year to the next? The final installment in our blog series on large language models (LLMs) and narrative disclosures explores exactly that question, showing how LLMs can be combined with structured XBRL data to compare narrative disclosures across reporting periods.
The blog demonstrates how tagging within the IFRS Accounting Taxonomy allows analysts to isolate comparable text sections, such as accounting policies or risk disclosures, and then use AI tools to highlight what has changed. Instead of needing to read through reports line by line, the approach allows users to quickly surface shifts in emphasis, policy updates, or new assumptions buried within lengthy narrative sections.
The result is a useful new lens on corporate reporting. By pairing structured data with AI analysis, analysts can quickly identify where disclosures have evolved, and where deeper investigation is needed. The full blog walks through the methodology and examples in detail, offering a glimpse of how structured reporting, combined with AI, could reshape how narrative disclosures are analysed.
Read the blog on our website here.
