Call for Papers

Topics of Interest

Financial, economic, and legal domains provide a distinctive testbed for the next generation of NLP and AI. Real-world financial information spans long documents, numerical tables, charts, news, regulatory filings, earnings calls, and rapidly evolving market signals; financial markets are multilingual, dynamic, and interconnected; and LLM agents, multi-agent systems, and world models are moving financial AI from static prediction toward interactive reasoning, simulation, and long-horizon decision making. We invite original research on these challenges across NLP, finance, economics, and law. Topics include, but are not limited to:

• Financial, economic, and legal language models and foundation models
• Numerical, tabular, long-context, and multimodal understanding and reasoning
• Multilingual, cross-lingual, low-resource, and domain-adaptive financial NLP
• LLM agents and multi-agent systems for financial and economic applications
• World models, market simulation, planning, memory, and sequential decision making
• Tool use, retrieval-augmented generation, and multi-source knowledge integration
• Knowledge graphs and structured representation learning for financial data
• Financial search, question answering, information extraction, and document intelligence
• Financial forecasting, event discovery, trading, portfolio management, and risk analysis
• Economic and market modeling, behavioral simulation, and decision support
• ESG analysis, sustainability assessment, and impact evaluation
• Financial regulation, compliance monitoring, legal reasoning, auditing, and regulatory reporting
• Synthetic data, data augmentation, privacy-preserving learning, and data quality
• Benchmarks and evaluation of financial models and agents
• Interpretability, fairness, privacy, security, safety, responsible AI, and governance
• Real-world systems, applications, and demonstrations from financial institutions and industry

Important Dates

// All deadlines will be 11:59 PM Anywhere on Earth (AoE, UTC-12)

How to Submit

FinNLP 2027 accepts two types of submissions:

1. Direct submissions, reviewed by the FinNLP program committee.
2. Pre-reviewed ARR commitments, with existing reviews from ACL Rolling Review.

Direct submissions will go through the FinNLP review process. ARR commitments should have already been reviewed through ARR and will be checked by the Program Chairs / Area Chairs for eligibility, scope, and final acceptance. The submission site and deadlines will be announced on this page.

Format

The ACL Template MUST be used for your submission(s). Accepted papers' proceedings will be published at the ACL Anthology.

• Long Paper: up to 8 pages, excluding references.
• Short Paper: up to 4 pages, excluding references.
• Demo Paper: up to 4 pages, excluding references.

Submission Policies

Anonymization

All direct submissions must be anonymized following the standard ACL guidelines:

• Do not include author names, affiliations, or acknowledgements in the submission PDF.
• Refer to your own prior work in the third person (e.g., "Smith et al. (2024) showed..." rather than "in our previous work...").
• Do not include institutional logos, identifiable URLs, or GitHub handles tied to the authors.
• Released code and data should be linked through anonymous repositories (e.g., anonymous.4open.science).
• Strip identifying PDF metadata before submission.

ARR-committed papers retain the anonymity status they had during ARR review.

Dual / multiple submission

• Papers must not be under review at another archival venue at the time of submission. If accepted, the same content cannot be submitted elsewhere.
• arXiv pre-prints are permitted, provided the submission remains anonymous in content.
• Papers committed via ARR must declare any other venues they are concurrently committed to.
• Concurrent submissions to other workshops, conferences, or journals must be disclosed at submission time.

Supplementary material

• Supplementary material (appendices, datasets, code) may be included in the appendix section (no page limit) or linked through anonymous repositories.
• Reviewers are not required to consult supplementary material; the core contribution must be evaluable from the main paper.
• For dataset and benchmark papers, we strongly encourage authors to include a data statement describing the data source, licensing, and ethical considerations.

Archival & non-archival

• Accepted papers are archival by default and published in the ACL Anthology proceedings.
• Authors may opt out of the proceedings at submission; such papers are designated non-archival, excluded from the official proceedings, and the authors may still present their work at the workshop.
• Non-archival papers — including work presented, under review, or scheduled to appear elsewhere (e.g., the ACL main conference or Findings) — may be submitted to other venues, subject to the other venue's dual-submission, anonymity, and prior-publication policies. Eligibility is ultimately determined by the other venue.

Review & Presentation

• Direct submissions will be reviewed through a double-blind review process by the FinNLP Program Chairs, Area Chairs, and Reviewers. There will be no author rebuttal period.
• Pre-reviewed ARR commitments will be checked by the Program Chairs / Area Chairs for eligibility, scope, and final acceptance.
• All direct submissions must be anonymized.
• At least one author of each accepted paper must register and present their work at FinNLP 2027. Presentation formats will follow ACL 2027's participation policy.

Open Research Policy

As part of ACL SIG-FinTech's reproducibility and transparency initiative, FinNLP continues its open research policy for accepted papers.

• Authors of accepted papers are strongly encouraged to publicly release their implementation code, model inputs, and model outputs to support reproducibility and future re-evaluation.
• For LLM-based approaches, authors are encouraged to release prompts, inference configurations, generated outputs, and evaluation artifacts associated with their experiments whenever possible.
• Exceptions may be granted in cases involving legal, licensing, privacy, commercial, or security constraints. Authors requesting exceptions should discuss their situation with the organizers.
• Presentation videos from the workshop will be uploaded to the ACL SIG-FinTech YouTube channel for public access and long-term educational use.

Contact

For questions about submissions or the workshop, contact aclsigfintech@gmail.com.