{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,23]],"date-time":"2026-10-23T16:26:09Z","timestamp":1792772769668,"version":"4.5.0"},"posted":{"date-parts":[[2026]]},"group-title":"SSRN","reference-count":69,"publisher":"Elsevier BV","license":[{"start":{"date-parts":[[2026,1,1]],"date-time":"2026-01-01T00:00:00Z","timestamp":1767225600000},"content-version":"unspecified","delay-in-days":0,"URL":"https:\/\/www.uspto.gov\/ip-policy\/copyright-policy\/copyright-basics"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"abstract":"<jats:p>This study examines the relationship between corporate artificial intelligence (AI) disclosures and firm-level idiosyncratic volatility. Using a validated large language model (LLM) benchmarked against two human coders, we analyze 65,485 AI-related sentences in the annual reports of Korean listed firms to measure promotional tone (AI Talk) and verifiable substantiation (AI Walk). On held-out data, the model&amp;apos;s agreement with the human Walk grades (weighted \u03ba = 0.81) approaches that between the coders (0.90). Across 3,831 firm-year observations from 2020 to 2024, firms with better-substantiated AI disclosures exhibit significantly lower idiosyncratic volatility in the following year, whereas promotional tone shows no robust association. The association holds across alternative volatility measures, additional controls, matching and a selection model, and it arises from differences between firms rather than from changes within firms. It is also distinct from AI capability: AI patents, a common proxy for such capability, are associated with higher volatility. Our findings suggest that the substantiation of AI disclosures is associated with lower firm-specific risk, and that a validated LLM can measure such substantiation where external data on AI activity are unavailable.<\/jats:p>","DOI":"10.2139\/ssrn.7661553","type":"posted-content","created":{"date-parts":[[2026,10,23]],"date-time":"2026-10-23T13:38:30Z","timestamp":1792762710000},"source":"Crossref","is-referenced-by-count":0,"title":["Talk or Walk? AI Disclosure and Firm Idiosyncratic Volatility: Evidence from Korea"],"prefix":"10.2139","author":[{"given":"Thi  Ngoc Anh","family":"Nguyen","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hail","family":"Jung","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"78","reference":[{"issue":"1","key":"ref1","doi-asserted-by":"crossref","first-page":"259","DOI":"10.1111\/j.1540-6261.2006.00836.x","article-title":"The Cross-Section of Volatility and Expected Returns","volume":"61","author":"A Ang","year":"2006","journal-title":"The Journal of Finance"},{"issue":"1","key":"ref2","doi-asserted-by":"crossref","first-page":"29","DOI":"10.17811\/ebl.13.1.2024.29-38","article-title":"The ChatGPT effect on AI-themed cryptocurrencies","volume":"13","author":"L Ante","year":"2024","journal-title":"Economics and Business Letters"},{"key":"ref3","doi-asserted-by":"crossref","DOI":"10.1016\/j.techfore.2024.123965","article-title":"Quantifying a firm's AI engagement: Constructing objective, datadriven, AI stock indices using 10-K filings","volume":"212","author":"L Ante","year":"2025","journal-title":"Technological Forecasting and Social Change"},{"key":"ref4","doi-asserted-by":"crossref","first-page":"10","DOI":"10.1016\/j.bir.2024.11.003","article-title":"Beyond polarity: How ESG sentiment influences idiosyncratic volatility in the Turkish stock market","volume":"24","author":"A Atak","year":"2024","journal-title":"Borsa Istanbul Review"},{"key":"ref5","doi-asserted-by":"crossref","DOI":"10.1016\/j.techfore.2022.122264","article-title":"Artificial intelligence and corporate innovation: A review and research agenda","volume":"188","author":"S Bahoo","year":"2023","journal-title":"Technological Forecasting and Social Change"},{"key":"ref6","article-title":"Artificial intelligence and stock price crash risk: Evidence from China: A pre-registered report","volume":"93","author":"X Bai","year":"2025","journal-title":"Pacific-Basin Finance Journal"},{"issue":"5","key":"ref7","doi-asserted-by":"crossref","first-page":"1661","DOI":"10.1111\/1475-679X.12542","volume":"62","author":"A C Baker","year":"2024","journal-title":"Diversity Washing. Journal of Accounting Research"},{"key":"ref8","article-title":"Signals or Smoke? The Determinants and Informativeness of Corporate Artificial Intelligence (AI) Disclosures","author":"J M Barrios","year":"2025","journal-title":"SSRN"},{"key":"ref9","first-page":"47","article-title":"Cheap talk and cherry-picking: What ClimateBert has to say on corporate climate risk disclosures","author":"J A Bingler","year":"2022","journal-title":"Finance Research Letters"},{"key":"ref10","doi-asserted-by":"crossref","DOI":"10.1016\/j.techsoc.2023.102440","article-title":"The Road to corporate sustainability: The importance of artificial intelligence","volume":"76","author":"P Chen","year":"2024","journal-title":"Technology in Society"},{"issue":"2","key":"ref11","doi-asserted-by":"crossref","first-page":"719","DOI":"10.1111\/j.1540-6261.2012.01729.x","article-title":"A Simple Way to Estimate Bid-Ask Spreads from Daily High and Low Prices","volume":"67","author":"S A Corwin","year":"2012","journal-title":"The Journal of Finance"},{"issue":"6","key":"ref12","doi-asserted-by":"crossref","first-page":"1431","DOI":"10.2307\/1913390","article-title":"Strategic Information Transmission","volume":"50","author":"V P Crawford","year":"1982","journal-title":"Econometrica"},{"issue":"9","key":"ref13","doi-asserted-by":"crossref","first-page":"7888","DOI":"10.1287\/mnsc.2023.03253","article-title":"ChatGPT for Textual Analysis? How to Use Generative LLMs in Accounting Research","volume":"71","author":"T De Kok","year":"2025","journal-title":"Management Science"},{"issue":"1","key":"ref14","doi-asserted-by":"crossref","first-page":"123","DOI":"10.2307\/2490910","article-title":"Disclosure of Nonproprietary Information","volume":"23","author":"R A Dye","year":"1985","journal-title":"Journal of Accounting Research"},{"key":"ref15","article-title":"Generative AI and Firm Values","author":"A L Eisfeldt","year":"2026","journal-title":"The Journal of Finance"},{"issue":"3-4","key":"ref16","doi-asserted-by":"crossref","first-page":"519","DOI":"10.1007\/s13162-025-00323-y","article-title":"AI washing: A conceptual exploration","volume":"15","author":"S Elhajjar","year":"2025","journal-title":"AMS Review"},{"issue":"1","key":"ref17","first-page":"90023","article-title":"Common risk factors in the returns on stocks and bonds","volume":"33","author":"E F Fama","year":"1993","journal-title":"Journal of Financial Economics"},{"key":"ref18","article-title":"Corporate Governance, Idiosyncratic Risk, and Information Flow","volume":"62","author":"M A Ferreira","year":"2007","journal-title":"Financial Investment Services and Capital Markets Act, Act No. 21324 \u00a7 Arts"},{"key":"ref19","article-title":"\uc2e0\uc0ac\uc5c5 \uc9c4\ud589\uc0c1\ud669 \uacf5\uc2dc \uc810\uac80 \ubc0f \uc0ac\uc5c5\uc9c4\ud589 \uc2e4\ud0dc\ubd84\uc11d \uacb0 \uacfc [Results of the review of new-business progress disclosures and analysis of business implementation","author":"Financial Supervisory","year":"2023","journal-title":"Korea"},{"key":"ref20","article-title":"\u300c\uc2e0\uc0ac\uc5c5 \uc9c4\ud589\uc0c1\ud669 \uacf5\uc2dc \uc810\uac80 \ubc0f \uc0ac\uc5c5\uc9c4\ud589 \uc2e4\ud0dc\ubd84\uc11d \uacb0\uacfc\u300d | \uacbd\uc81c\uc815\ucc45\uc790\ub8cc","author":"Financial Supervisory","year":"2024","journal-title":"KDI \uacbd\uc81c\uad50\uc721"},{"key":"ref21","author":"Fnguide","year":"2025","journal-title":"\uc81c\uacf5 \ub370\uc774\ud130 \uc548\ub0b4"},{"issue":"1","key":"ref22","doi-asserted-by":"crossref","first-page":"24","DOI":"10.1016\/j.jfineco.2008.02.003","article-title":"Idiosyncratic risk and the cross-section of expected stock returns","volume":"91","author":"F Fu","year":"2009","journal-title":"Journal of Financial Economics"},{"key":"ref23","article-title":"AI washing under media pressure: Evidence from China","volume":"214","author":"M Fu","year":"2026","journal-title":"Journal of Business Research"},{"issue":"3","key":"ref24","doi-asserted-by":"crossref","first-page":"461","DOI":"10.1086\/466995","article-title":"The Informational Role of Warranties and Private Disclosure about Product Quality","volume":"24","author":"S J Grossman","year":"1981","journal-title":"The Journal of Law and Economics"},{"issue":"7","key":"ref25","doi-asserted-by":"crossref","first-page":"1720","DOI":"10.1108\/IJOEM-08-2020-0933","article-title":"Herding and loss aversion in stock markets: Mediating role of fear of missing out (FOMO) in retail investors","volume":"17","author":"S Gupta","year":"2021","journal-title":"International Journal of Emerging Markets"},{"issue":"1","key":"ref26","doi-asserted-by":"crossref","first-page":"25","DOI":"10.1093\/pan\/mpr025","article-title":"Entropy Balancing for Causal Effects: A Multivariate Reweighting Method to Produce Balanced Samples in Observational Studies","volume":"20","author":"J Hainmueller","year":"2012","journal-title":"Political Analysis"},{"issue":"1","key":"ref27","doi-asserted-by":"crossref","first-page":"153","DOI":"10.2307\/1912352","article-title":"Sample Selection Bias as a Specification Error","volume":"47","author":"J J Heckman","year":"1979","journal-title":"Econometrica"},{"issue":"3","key":"ref28","doi-asserted-by":"crossref","first-page":"632","DOI":"10.1016\/j.jfineco.2012.09.011","article-title":"Innovative efficiency and stock returns","volume":"107","author":"D Hirshleifer","year":"2013","journal-title":"Journal of Financial Economics"},{"issue":"4","key":"ref29","doi-asserted-by":"crossref","first-page":"1005","DOI":"10.1007\/s11142-016-9371-1","article-title":"The benefits of specific risk-factor disclosures","volume":"21","author":"O.-K Hope","year":"2016","journal-title":"Review of Accounting Studies"},{"key":"ref30","author":"K Hu","year":"2023","journal-title":"ChatGPT sets record for fastest-growing user base-Analyst note"},{"key":"ref31","doi-asserted-by":"crossref","DOI":"10.1016\/j.ememar.2023.101045","article-title":"Corporate digital transformation and idiosyncratic risk: Based on corporate governance perspective","volume":"56","author":"H Huang","year":"2023","journal-title":"Emerging Markets Review"},{"issue":"2","key":"ref32","doi-asserted-by":"crossref","first-page":"596","DOI":"10.1111\/joms.13127","article-title":"Industry Exposure to Artificial Intelligence, Board Network Heterogeneity, and Firm Idiosyncratic Risk","volume":"63","author":"K Hudson","year":"2026","journal-title":"Journal of Management Studies"},{"key":"ref33","doi-asserted-by":"crossref","DOI":"10.1016\/j.frl.2023.104052","article-title":"Carbon productivity and volatility","volume":"56","author":"H Jung","year":"2023","journal-title":"Finance Research Letters"},{"key":"ref34","article-title":"Klarna AI assistant handles two-thirds of customer service chats in its first month","author":"Klarna","year":"2024","journal-title":"Klarna"},{"key":"ref35","author":"B Li","year":"2024","journal-title":"AI Washing"},{"key":"ref36","article-title":"The impact of AI washing on enterprises' access to bank loans: From the perspective of external governance","volume":"98","author":"W Liu","year":"2026","journal-title":"Finance Research Letters"},{"key":"ref37","article-title":"Exploring the Relationship Between Human-Centric AI and Firm Idiosyncratic Risks","author":"Z R Liu","year":"2026","journal-title":"Information Systems Frontiers"},{"key":"ref38","article-title":"All that glitters? Corporate AI washing and stock price crash risk","volume":"99","author":"Z Liu","year":"2026","journal-title":"Finance Research Letters"},{"issue":"0","key":"ref39","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1080\/13504851.2026.2745636","article-title":"When AI narratives shake the market: AI-washing and idiosyncratic stock return volatility in China","volume":"0","author":"D Luo","year":"2026","journal-title":"Applied Economics Letters"},{"issue":"2","key":"ref40","doi-asserted-by":"crossref","first-page":"380","DOI":"10.2307\/3003562","article-title":"Good News and Bad News: Representation Theorems and Applications","volume":"12","author":"P R Milgrom","year":"1981","journal-title":"The Bell Journal of Economics"},{"key":"ref41","article-title":"2025 ingongjineung saneop siltae josa [2025 artificial intelligence industry survey","year":"2026","journal-title":"Ministry of Science and ICT, & Software Policy & Research Institute"},{"issue":"1","key":"ref42","doi-asserted-by":"crossref","first-page":"69","DOI":"10.2307\/1913646","article-title":"On the Pooling of Time Series and Cross Section Data","volume":"46","author":"Y Mundlak","year":"1978","journal-title":"Econometrica"},{"issue":"2","key":"ref43","doi-asserted-by":"crossref","first-page":"187","DOI":"10.1080\/07350015.2016.1227711","article-title":"Unobservable Selection and Coefficient Stability: Theory and Evidence","volume":"37","author":"E Oster","year":"2019","journal-title":"Journal of Business & Economic Statistics"},{"issue":"6","key":"ref44","doi-asserted-by":"crossref","first-page":"696","DOI":"10.1177\/14707853251379285","article-title":"Responsible AI in Marketing: AI Booing and AI Washing Cycle of AI Mistrust","volume":"67","author":"S Ozturkcan","year":"2025","journal-title":"International Journal of Market Research"},{"issue":"3","key":"ref45","doi-asserted-by":"crossref","first-page":"1113","DOI":"10.1111\/j.1540-6261.2012.01743.x","article-title":"Investment, Idiosyncratic Risk, and Ownership","volume":"67","author":"V Panousi","year":"2012","journal-title":"The Journal of Finance"},{"key":"ref46","author":"J Park","year":"2026","journal-title":"ESG Disclosure Quality as Organizational Information Processing: Comparing Human Coding, Rule-Based Automation, and LLM Semantic Scoring. Business Strategy and the Environment, bse.71277"},{"issue":"4","key":"ref47","doi-asserted-by":"crossref","first-page":"1","DOI":"10.17549\/gbfr.2018.23.4.1","article-title":"The Comparison of Trading Costs between Two Markets: KOSPI and KOSDAQ","volume":"23","author":"K B Park","year":"2018","journal-title":"Global Business Finance Review"},{"issue":"5","key":"ref48","doi-asserted-by":"crossref","first-page":"1749","DOI":"10.1111\/1540-6261.00587","article-title":"Stock Valuation and Learning about Profitability","volume":"58","author":"\u013d P\u00e1stor","year":"2003","journal-title":"The Journal of Finance"},{"key":"ref49","doi-asserted-by":"crossref","DOI":"10.1016\/j.eneco.2023.107053","article-title":"The impact of carbon disclosure and carbon emissions intensity on firms' idiosyncratic volatility","volume":"128","author":"K Perera","year":"2023","journal-title":"Energy Economics"},{"issue":"1","key":"ref50","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1016\/j.jacceco.2010.06.001","article-title":"Financial reporting quality and idiosyncratic return volatility","volume":"51","author":"S Rajgopal","year":"2011","journal-title":"Journal of Accounting and Economics"},{"key":"ref51","article-title":"Is more not always better? Untangling the impact of ESG disclosure transparency on idiosyncratic risk","volume":"373","author":"J Ren","year":"2025","journal-title":"Journal of Environmental Management"},{"key":"ref52","article-title":"Does geopolitical risk affect firms' idiosyncratic volatility? Evidence from China","volume":"90","author":"X Ren","year":"2023","journal-title":"International Review of Financial Analysis"},{"key":"ref53","article-title":"Climate policy uncertainty and idiosyncratic volatility: Evidence from the non-financial listed Chinese firms","volume":"5","author":"X Ren","year":"2023","journal-title":"Journal of Climate Finance"},{"issue":"1","key":"ref54","doi-asserted-by":"crossref","first-page":"41","DOI":"10.1093\/biomet\/70.1.41","article-title":"The central role of the propensity score in observational studies for causal effects","volume":"70","author":"P R Rosenbaum","year":"1983","journal-title":"Biometrika"},{"key":"ref55","author":"Sec","year":"2024","journal-title":"SEC Announces Enforcement Results for Fiscal Year 2024"},{"key":"ref56","author":"Sec","year":"2024","journal-title":"SEC Charges Two Investment Advisers with Making False and Misleading Statements About Their Use of Artificial Intelligence. U.S. Securities and Exchange Commission"},{"key":"ref57","author":"S Shibu","year":"2025","journal-title":"Klarna Is Hiring Customer Service Agents After AI Couldn't Cut It on Calls, According to the Company's CEO"},{"key":"ref58","doi-asserted-by":"crossref","DOI":"10.1016\/j.frl.2026.109684","article-title":"AI washing: Strategic disclosure and backlash","volume":"95","author":"X Song","year":"2026","journal-title":"Finance Research Letters"},{"issue":"3","key":"ref59","doi-asserted-by":"crossref","first-page":"355","DOI":"10.2307\/1882010","article-title":"Job Market Signaling","volume":"87","author":"M Spence","year":"1973","journal-title":"The Quarterly Journal of Economics"},{"key":"ref60","first-page":"32","volume":"79","author":"J C Stein","year":"1989","journal-title":"Cheap Talk and the Fed: A Theory of Imprecise Policy Announcements"},{"key":"ref61","article-title":"Unveiling AI washing: Bridging corporate technological gaps through a cognitive dissonance lens","volume":"225","author":"Z Sun","year":"2026","journal-title":"Technological Forecasting and Social Change"},{"key":"ref62","article-title":"Greener pastures, steadier returns: ESG ratings and idiosyncratic risk management","volume":"100","author":"M Tao","year":"2025","journal-title":"International Review of Economics & Finance"},{"key":"ref63","doi-asserted-by":"crossref","first-page":"179","DOI":"10.1016\/0165-4101(83)90011-3","article-title":"Discretionary disclosure","volume":"5","author":"R E Verrecchia","year":"1983","journal-title":"Journal of Accounting and Economics"},{"key":"ref64","article-title":"Policy-induced AI Talk-Walk gaps and stock price crash risk: Evidence from China's AI pilot zones","volume":"164","author":"C Wu","year":"2026","journal-title":"Economic Modelling"},{"issue":"7","key":"ref65","article-title":"Beyond AI Narratives: AI Washing and Organizational Resilience","volume":"14","author":"Y Xia","year":"2026","journal-title":"Systems"},{"key":"ref66","author":"D Zachary","year":"2024","journal-title":"AI Washing\" -The New Deceptive Marketing Technique. The Pomerantz Monitor (Pomerantz LLP)"},{"issue":"2","key":"ref67","doi-asserted-by":"crossref","first-page":"503","DOI":"10.1111\/jfir.12424","article-title":"Artificial intelligence innovation and stock price crash risk","volume":"48","author":"J Zhang","year":"2025","journal-title":"Journal of Financial Research"},{"issue":"1","key":"ref68","doi-asserted-by":"crossref","first-page":"237","DOI":"10.1162\/coli_a_00502","article-title":"Can Large Language Models Transform Computational Social Science?","volume":"50","author":"C Ziems","year":"2024","journal-title":"Computational Linguistics"},{"key":"ref69","author":"Columns","journal-title":") and (3) estimate the joint model separately for the two periods (1,087 and 2,744 firm-years). Columns (4)-(6): column (4) adds KOSDAQ and AI Walk \u00d7 KOSDAQ; columns (5) and (6) estimate the joint model separately for KOSPI (1,138) and KOSDAQ (2,693) firm-years. AI Walk and AI Talk are standardized; controls are defined in Appendix A. Coefficients are multiplied by 100 (percentage points of daily volatility), except lagged IVOL, which is in percent like the dependent variable; constants are not reported. Standard errors clustered by firm in parentheses"}],"container-title":[],"original-title":[],"deposited":{"date-parts":[[2026,10,23]],"date-time":"2026-10-23T13:40:49Z","timestamp":1792762849000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.ssrn.com\/abstract=7661553"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026]]},"references-count":69,"URL":"https:\/\/doi.org\/10.2139\/ssrn.7661553","relation":{},"subject":[],"published":{"date-parts":[[2026]]},"subtype":"preprint"}}