Unlocking the Potential of DH+AI: Opportunities, Challenges, and Recommendations

DH+AI as a Transformative and Emerging Field

The intersection of Digital Humanities (DH) and Artificial Intelligence (AI) is rapidly emerging as a crucial field, offering significant opportunities for both the humanities and social sciences (HSS) and AI research itself. DH+AI not only enhances HSS methodologies — enabling new forms of data analysis, historical research, and knowledge representation — but also contributes to AI development by providing critical insights into ethical considerations, interpretability, and human-centred applications.

Researchers in HSS increasingly leverage AI to model complex social phenomena, analyze vast textual corpora, and develop predictive frameworks for social, political, and cultural trends. At the same time, HSS disciplines critically examine AI’s broader societal impacts, from disinformation and biases in machine learning models to AI’s influence on democracy, governance, and digital culture. This dual role makes HSS indispensable in shaping responsible AI development and deployment.

However, despite its transformative potential, DH+AI research remains structurally and methodologically fragmented. Many initiatives remain isolated within specific academic disciplines, research programmes, or funding schemes. To fully unlock the potential of this field, several key challenges must be addressed, particularly in ethical, methodological, and institutional domains.

Challenges in Advancing DH+AI Research

Ethical and Equity Considerations

The sustainability of AI in HSS depends on strong ethical frameworks that address issues such as intellectual property, data privacy, and equitable access to resources. While public debates often focus on privacy and algorithmic fairness, DH+AI research must also consider:

• Sensitive and contested data
  Many HSS disciplines handle complex and politically charged datasets, such as historical records tied to conflicting memories. AI’s application to these datasets requires careful methodological revision to prevent misrepresentation or misuse
• Preserving small but critical datasets
  While AI research often prioritizes large-scale data, HSS relies on small, unique, and contextually rich datasets. Ensuring their digitization, accessibility, and long-term preservation is vital for cultural memory and scholarly inquiry.
• Open science and AI governance
  AI’s growing role in knowledge production necessitates democratic oversight, particularly in cases where AI models influence public discourse, policymaking, and historical narratives. Greater transparency and regulatory measures are needed to prevent biases and reinforce trust in AI-data-driven research.

Methodological and Interdisciplinary Challenges

DH and HSS disciplines are characterized by interpretive, qualitative, and historically grounded approaches, which contrast with AI’s quantitative and algorithmic nature. This epistemological gap presents several challenges:

• Standardizing data representation and analysis
  While AI requires structured and standardized datasets, many HSS disciplines work with highly heterogeneous, non-standardized sources. Developing interoperable frameworks for knowledge representation is critical for meaningful AI integration.
• Bridging qualitative and quantitative methodologies
  AI’s reliance on computational methods risks oversimplifying complex humanistic and social phenomena. More interdisciplinary collaboration is needed to ensure that AI tools accurately capture the plurality of perspectives in HSS research.
• Ensuring transparency and explainability
  Machine learning models, especially large language models (LLMs), often function as ‘black boxes,’ making it difficult for researchers to critically assess their outputs. Enhancing AI transparency and explainability is essential to maintaining academic rigor and public trust.

Institutional and Structural Barriers

Despite growing recognition of DH+AI’s significance, institutional challenges hinder its development:

• Fragmentation in research funding
  In European frameworks such as Horizon Europe (HE), DH+AI topics are scattered across multiple sub-areas and funding calls, making it difficult for researchers to position their work effectively. This also complicates evaluation processes, as panel reviewers often lack interdisciplinary expertise.
• Disparities in funding allocation
  AI research, particularly in STEM fields, receives significantly more funding than HSS, limiting opportunities for genuine interdisciplinary collaboration. HSS-led initiatives struggle to attract top AI researchers, leading to a one-sided dynamic where AI projects merely extract HSS data without engaging in substantive collaboration.
• Lack of research infrastructure
  While AI benefits from large-scale computing facilities, HSS research often lacks equivalent infrastructure for digital scholarship. 

Recommendations for Advancing DH+AI Research

To address these challenges and foster the integration of AI and HSS, we propose the following strategic recommendations:

• Strengthening Research and Funding Structures
  Encourage interdisciplinary funding mechanisms
  Horizon Europe and other funding bodies should create dedicated DH+AI programmes, reducing fragmentation and ensuring sustained support for interdisciplinary initiatives.
• Introduce Proof-of-Concept (PoC) calls
  Many DH research outputs—such as digital archives, methodologies, and ethical guidelines—are not traditionally ‘marketable’ but hold significant societal value. PoC calls could help translate these research findings into practical applications, from policy recommendations to AI-driven tools for cultural heritage.
• Enhance HSS representation in AI research clusters
  DH and HSS scholars should have a stronger presence in major AI research initiatives, ensuring that ethical, historical, and societal perspectives shape AI development from the outset.

Building Institutional and International Cooperation

• Strengthen international collaborations
  Institutions such as CNR, CNRS, and CSIC have a strong tradition of cooperation, which should be reinforced through renewed framework agreements, joint research programmes, and funding schemes that facilitate interdisciplinary exchange.
• Develop shared research infrastructure
  National research infrastructures should be made accessible across institutions to support collaborative DH+AI projects. Key HSS datasets should be recognized as strategic research assets, ensuring their long-term sustainability and accessibility and an effort should be made as well in providing the HSS team with computing power and digital engineers.
• Establish a dedicated interdisciplinary research programme
  Following models like the CNRS’s International Research Laboratories, a new programme should be created to fund joint DH+AI projects that align with key societal challenges.

Promoting Knowledge Exchange and Capacity Building

• Support interdisciplinary training
  AI researchers should receive training in HSS methodologies, while HSS scholars should develop computational literacy to critically engage with AI tools. Specialized educational programmes should equip future researchers with the necessary skills to navigate this interdisciplinary space.
• Foster open science and data sharing
  A trilateral forum between CNR, CNRS, and CSIC should be established to address open science, data transparency, and ethical AI applications in HSS.
• Establish cross institutions scientific committee
  This committee would maintain a scientific watch on this dynamic field, organise a biannual DH+AI conference and be a strategic consulting body for the institutions.

Conclusion

The integration of AI and Digital Humanities represents a critical frontier for both HSS and AI research, offering new methodological possibilities while raising important societal and ethical questions. However, several structural, methodological, and institutional challenges must be addressed to fully realize the potential of DH+AI research.

By fostering interdisciplinary funding mechanisms, strengthening international collaboration, and enhancing knowledge exchange, we can build a more cohesive and impactful DH+AI research ecosystem. AI’s role in shaping knowledge, culture, and policy will continue to expand, making it essential that the humanities and social sciences play a central role in guiding its development. Through strategic investment and institutional support, DH+AI can contribute not only to the advancement of HSS but also to the broader goal of ensuring AI’s responsible and equitable integration into society.

Gianluca Fasano (CNR-ISTC), Cristina Marras (CNR-ILIESI), Emilio Sanfilippo (CNR-ISTC), Monica Brînzei (CNRS-IRHT), Mathieu Husson (CNRS-LTE), Paola Tubaro (CNRS-CREST), Patricia Martín-Rodilla (CSIC-IEGPS); Daniel Riaño (CSIC-CCHS), Ana Gómez Rabal (CSIC-IMF)