Fast-paced business landscape nowadays has rendered the integrity of financial reporting pivotal for stakeholders ranging from investors to board members. The concept of Internal Control over Financial Reporting (ICFR), therefore, has gained prominence, driven by evolving guidance and regulations. Public companies navigate the fine line between compliance and safeguarding proprietary information. This research examines linguistic complexity in 15,000 ICFR disclosures from over 6,000 US public companies across industries and over years, by leveraging Natural Language Processing (NLP) techniques. Furthermore, regression modeling is employed to reveal a significant relationship between linguistic complexity and material weakness in internal controls. Notably, word complexity (measured by syllable count of the disclosure) and text complexity (measured by the percentage of boilerplates, redundant words and difficult words in the disclosure) emerge as key explanatory determinants, with variations observed across industries and over time. These findings offer actionable insights for auditors and investors in evaluating public disclosures while informing policymakers and regulators of necessary linguistic elements that could be used for potential adjustments on mandatory disclosure regulations.