IOP Publishing has launched a new diamond open-access academic journal, AI Horizons, created to publish peer-reviewed research on artificial intelligence applications in science and engineering. Announced in an October 19, 2026 report by Physics World, the publication is a joint venture with the Eastern Institute of Technology (EIT) in Ningbo, the Zhejiang Association for Science and Technology, and the World Association of Young Scientists. By operating under a diamond open-access model, the journal makes all published papers immediately free to readers while charging no article-processing fees to authors, with partner sponsors covering the underlying publication costs.

Bridging Physical Models and Computational Intelligence

The editorial direction of AI Horizons is led by editor-in-chief Shiyi Chen, who currently serves as president of the Eastern Institute of Technology. Chen brings an extensive background in physics and engineering, having previously worked at Johns Hopkins University in the United States, Peking University, and the Southern University of Science and Technology. Historically focused on turbulence and computational fluid dynamics, Chen’s recent research has expanded into artificial intelligence for industrial and engineering applications.

In outlining the vision for the publication, Chen emphasized that while machine learning is altering how researchers process data and build scientific models, the actual scientific value depends on how well those results hold up to empirical testing. The journal aims to highlight research that pairs computational techniques with established physical principles and direct comparisons against physical experiments or observations.

One specific area of focus is the integration of machine learning into complex physical simulations, such as fluid turbulence, which requires modeling interactions across multiple physical scales. Combining neural networks with traditional physical models can improve computer-aided engineering and industrial software, allowing engineers to evaluate structural and mechanical designs more efficiently than conventional computation allows.

Technical Scope and Cross-Disciplinary Peer Review

To differentiate itself from generalist machine learning conferences and journals, AI Horizons centers its remit on four primary domains: AI for science and engineering, spatiotemporal intelligence, embodied AI, and quantum intelligence. Alongside applied research that resolves difficult scientific questions or makes previously impractical engineering calculations feasible, the editorial board welcomes papers detailing fundamental advances in underlying AI methodologies.

However, the publication sets specific expectations for reproducibility and transparency. According to Chen, submitting authors must clearly demonstrate what their methods make possible, provide rigorous evidence backing their claims, and explicitly identify where their computational methods reach their operational limits. The goal is to provide a repository of verified tools that specialists across disparate scientific fields can test and build upon.

Evaluating papers that sit at the intersection of computer science and physical engineering often presents a bottleneck in traditional academic publishing. To address this, AI Horizons plans to structure its peer-review process by pairing reviewers who understand the core algorithmic methods with domain experts who understand the specific scientific problem being addressed. The editorial team also plans to commission cross-disciplinary review articles and perspective pieces to help researchers translate computational techniques from one scientific domain to another.

Commercial Context and Open-Access Impact

The debut of a dedicated, fee-free academic venue arrives as artificial intelligence continues its rapid expansion across both foundational research and commercial enterprise. Outside of academia, commercial deployment of automated systems is accelerating globally. For example, the Türkiye Artificial Intelligence Initiative (TRAI) reported in its October 2026 update that the number of tracked AI startups in the country rose from 482 in April 2026 to 504, according to Hürriyet Daily News. That growth was led by agentic AI startups, LegalTech, HR Tech, and GPU infrastructure providers.

Similarly, enterprise adoption in the United States continues to focus on operational workflows. On October 19, Huntington National Bank debuted its 1866 Studio, an operating system built on Google’s Gemini platform to provide small businesses with agentic AI workflows for financial, operational, and management tasks, according to PYMNTS.com.

While commercial sectors focus heavily on software automation and business workflows, scientific researchers face different hurdles—chiefly the high costs of academic publishing and the need for domain-specific validation of scientific machine learning models. By removing financial barriers for both authors and readers, AI Horizons offers university labs and industrial research groups an accessible venue to share applied engineering breakthroughs.

Frequently Asked Questions

What is the diamond open-access model used by AI Horizons?

Under a diamond open-access framework, published research articles are made completely free for the public and the research community to read, download, and distribute. Unlike gold open-access journals, which require authors or their institutions to pay substantial article-processing charges (APCs), diamond open-access journals charge nothing to authors. Instead, the publishing costs are underwritten by institutional sponsors and partner organizations.

Which organizations are partnering to publish the journal?

AI Horizons is published by IOP Publishing in partnership with the Eastern Institute of Technology (EIT) in Ningbo, the Zhejiang Association for Science and Technology, and the World Association of Young Scientists.

What types of research does AI Horizons accept?

The journal focuses on research that advances scientific understanding and engineering practice. Its specific scope encompasses AI for science and engineering, spatiotemporal intelligence, embodied AI, and quantum intelligence, as well as fundamental advances in AI methodology that clearly define both their capabilities and technical limitations.

Sources

Read More