Reasoning for Safety in Artificial Intelligence: Optimisation, Multi-Agent Systems, and Trustworthy Inference
A special issue of Big Data and Cognitive Computing (ISSN 2504-2289).
Deadline for manuscript submissions: 3 December 2026 | Viewed by 36
Special Issue Editor
Special Issue Information
Dear Colleagues,
Ensuring that AI systems reason reliable, transparent, and in alignment with human objectives has become a central challenge in contemporary AI safety research. While language and vision models have impressive generalisation ability, their reasoning processes often rely on shallow heuristics, pattern matching, or sycophantic tendencies rather than genuine inference. Recent findings emphasise the importance of structured reasoning methods—optimisation-based techniques, multi-agent debate, agentic pipelines, and reflective or self-verifying procedures—to support robust, auditable, and trustworthy decision-making.
This Special Issue focuses on Reasoning for Safety in AI, aiming to consolidate emerging advances and stimulate novel research directions. BDCC is a natural venue for this topic, given its emphasis on data, computation, and complex systems that increasingly rely on autonomous reasoning agents.
We welcome contributions that investigate how formal, semi-formal, or hybrid reasoning mechanisms can enhance safety, robustness, and oversight in AI systems.
Topics of interest include, but are not limited to, novel reasoning algorithms, agentic architectures, debate-based oversight, optimisation-guided inference, and applied safety frameworks in high-stakes domains, as well as the following:
- Optimisation-based reasoning (GRPO, trajectory optimisation, constraint-guided inference);
- Agentic and multi-agent reasoning for oversight (debate, self-critique, weak-to-strong supervision);
- Benchmarking and evaluation of reasoning safety in LLMs and MLLMs;
- Argumentation and dialectical and symbolic reasoning for verifiable safety;
- Reasoning under uncertainty, reliability of chain-of-thought, and mitigation of pseudo-reasoning;
- Trustworthy multi-modal inference and cross-lingual safety;
- AI safety frameworks using reasoning (risk detection, audit chains, compliance reasoning, real-world alignment);
- Interpretability methods for explaining reasoning processes;
- Applications of reasoning-enhanced AI in healthcare, finance, law, education, and public governance.
We look forward to receiving your contributions.
Dr. Leonardo Ranaldi
Guest Editor
Ms. Giulia Pucci
Guest Editor Assistant
Affiliation: Department of Computing Science, University of Aberdeen, AB10 Aberdeen, Scotland, UK
Email: giulia.pucci@abdn.ac.uk
Interests: AI safety; human-machine interaction; AI adaption
Website: https://www.linkedin.com/in/giuliapucci1202/?locale=en_US
Manuscript Submission Information
Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 250 words) can be sent to the Editorial Office for assessment.
Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Big Data and Cognitive Computing is an international peer-reviewed open access monthly journal published by MDPI.
Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 1800 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.
Keywords
- reasoning in AI
- multi-agent systems
- agentic AI
- safety and oversight
- optimisation-based reasoning
- dialectical argumentation
- trustworthy AI
- weak-to-strong supervision
- reasoning evaluation
- explainability and verification
Benefits of Publishing in a Special Issue
- Ease of navigation: Grouping papers by topic helps scholars navigate broad scope journals more efficiently.
- Greater discoverability: Special Issues support the reach and impact of scientific research. Articles in Special Issues are more discoverable and cited more frequently.
- Expansion of research network: Special Issues facilitate connections among authors, fostering scientific collaborations.
- External promotion: Articles in Special Issues are often promoted through the journal's social media, increasing their visibility.
- Reprint: MDPI Books provides the opportunity to republish successful Special Issues in book format, both online and in print.
Further information on MDPI's Special Issue policies can be found here.
