Bedrock AI uses state of the art machine-learned algorithms to extract disclosures and assess risk from financial text. Our models pick-up on key environmental, product, and governance issues. We provide risk score for SEC registrants issuing 10-Ks, 10-Qs, or 20-Fs. Bedrock AI uses state of the art machine-learned algorithms to extract impactful disclosures and red-flags from financial text. This data-set covers disclosures that would be identified by forensic accountants. Former CFRA analysts have verified our results. Includes 10-K, 10-Q, and 20-F coverage. Most public company data is unstructured and textual. This data is critical to understanding company performance and risk (quality of earnings, governance, and management integrity), but is typically underused because of the difficulty of extracting impactful text. Our models focus on extracting disclosures correlated with high impact downside risk. This data-set includes disclosures on accounting and governance choices, product defects, regulatory issues, customer/market reliance and much more. We cover all SEC registrants issuing 10-Ks, 10-Qs, or 20-Fs. Extraction is complete within minutes of filings becoming available through EDGAR.