{"id":5593,"date":"2026-01-19T09:40:27","date_gmt":"2026-01-19T09:40:27","guid":{"rendered":"https:\/\/conferences.sigappfr.org\/wims2026\/?post_type=speaker&p=5593"},"modified":"2026-04-22T19:29:16","modified_gmt":"2026-04-22T19:29:16","slug":"sophia-ananiadou","status":"publish","type":"speaker","link":"https:\/\/conferences.sigappfr.org\/wims2026\/speaker\/sophia-ananiadou\/","title":{"rendered":"Sophia Ananiadou"},"content":{"rendered":"
\n
\"\"<\/figure>\n<\/div>\n\n\n

Sophia Ananiadou<\/h1>\n\n\n\n

Professor at National Centre for Text Mining, The University of Manchester, UK<\/h4>\n\n\n\n

Sophia Ananiadou is Professor of Computer Science at the University of Manchester, Director of the National Centre for Text Mining, Deputy Director of the Pankhurst Institute, and an ELLIS Fellow. She also holds an adjunct professorship at the School of Artificial Intelligence at Wuhan University and is a lead researcher at the Archimedes Research Unit in Greece.<\/p>\n\n\n\n

Her research focuses on interpretable, robust, and trustworthy NLP for extracting structured knowledge from large-scale unstructured text. A pioneer of biomedical text mining, she has developed foundational tools and platforms for modelling complex relationships to support drug repurposing, clinical research, and scientific discovery.<\/p>\n\n\n\n

Her recent work extends to agentic AI, culturally aware and emotionally robust large language models, and cross-domain misinformation detection. She also leads internationally recognised research in mental health NLP, developing interpretable models for detecting depression, psychosis, and cognitive decline. Her broader contributions span automatic summarisation, evidence-based medicine, systematic reviews, and financial NLP, including systems for evidence synthesis, risk analysis, and regulatory compliance.<\/p>\n\n\n\n

Title: Agentic AI in Finance: Risk, Reasoning, and Trust<\/h2>\n\n\n\n

Abstract<\/strong>: Financial AI is rapidly moving beyond prediction and analysis toward systems that influence and execute real-world decisions. This talk explores how large language models (LLMs) and emerging agentic AI systems are reshaping financial behaviour, from interpreting complex information and assessing risk to acting within markets.<\/p>\n\n\n\n

Drawing on our recent work, including FinBen for financial reasoning, agent-based trading systems, and FINCHAIN for verifiable multi-step reasoning, I examine what these systems can reliably do, as well as where they fail. A central focus is financial misinformation: an area where current models exhibit systematic weaknesses that are not yet well captured by standard evaluations.<\/p>\n\n\n\n

While capabilities are improving, key weaknesses remain. Models struggle with reference-free detection of plausible misinformation, exhibit context- and scenario-dependent variability in judgment, and show uneven performance across languages and financial domains.<\/p>\n\n\n\n

I argue that robust governance in high-stakes settings requires benchmarking, transparency, and open, scenario-aware evaluation. As AI systems move from describing markets to acting within them, ensuring reliability under realistic conditions becomes the central challenge.<\/p>\n\n\n\n

<\/p>\n\n\n\n

<\/p>\n","protected":false},"excerpt":{"rendered":"

Sophia Ananiadou Professor at National Centre for Text Mining, The University of Manchester, UK Sophia Ananiadou is Professor of Computer Science at the University of Manchester, Director of the National Centre for Text Mining, Deputy Director of the Pankhurst Institute,…<\/p>\n","protected":false},"author":3,"featured_media":5594,"comment_status":"open","ping_status":"closed","template":"","meta":{"footnotes":""},"speaker-category":[141],"speaker-tags":[148,109],"location":[192,191],"class_list":["post-5593","speaker","type-speaker","status-publish","has-post-thumbnail","hentry","speaker-category-keynote-speaker","speaker-tags-keynotes","speaker-tags-speaker","location-manchester-united-kingdom","location-united-kingdom"],"_links":{"self":[{"href":"https:\/\/conferences.sigappfr.org\/wims2026\/wp-json\/wp\/v2\/speaker\/5593","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/conferences.sigappfr.org\/wims2026\/wp-json\/wp\/v2\/speaker"}],"about":[{"href":"https:\/\/conferences.sigappfr.org\/wims2026\/wp-json\/wp\/v2\/types\/speaker"}],"author":[{"embeddable":true,"href":"https:\/\/conferences.sigappfr.org\/wims2026\/wp-json\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/conferences.sigappfr.org\/wims2026\/wp-json\/wp\/v2\/comments?post=5593"}],"version-history":[{"count":2,"href":"https:\/\/conferences.sigappfr.org\/wims2026\/wp-json\/wp\/v2\/speaker\/5593\/revisions"}],"predecessor-version":[{"id":5604,"href":"https:\/\/conferences.sigappfr.org\/wims2026\/wp-json\/wp\/v2\/speaker\/5593\/revisions\/5604"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/conferences.sigappfr.org\/wims2026\/wp-json\/wp\/v2\/media\/5594"}],"wp:attachment":[{"href":"https:\/\/conferences.sigappfr.org\/wims2026\/wp-json\/wp\/v2\/media?parent=5593"}],"wp:term":[{"taxonomy":"speaker-category","embeddable":true,"href":"https:\/\/conferences.sigappfr.org\/wims2026\/wp-json\/wp\/v2\/speaker-category?post=5593"},{"taxonomy":"speaker-tags","embeddable":true,"href":"https:\/\/conferences.sigappfr.org\/wims2026\/wp-json\/wp\/v2\/speaker-tags?post=5593"},{"taxonomy":"location","embeddable":true,"href":"https:\/\/conferences.sigappfr.org\/wims2026\/wp-json\/wp\/v2\/location?post=5593"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}