Transforming Global Animal Health Information Systems: Towards Trusted, AI-Assisted Reporting

After nearly two decades working with WAHIS and leading its development since 2015, I have seen the platform evolve alongside shifting disease landscapes, rising expectations and surging data volumes. While WOAH is not yet deploying AI in WAHIS, it is exploring the conditions for its responsible use. 

WAHIS holds more than 20 years of official, verified global animal health data submitted by national Veterinary Authorities. This continuity provides a structured, longitudinal view of disease reporting, making it a valuable asset that many AI initiatives lack. Whereas fragmented or incomplete datasets can produce biased predictions or unreliable outputs, WAHIS offers a relatively stable foundation for developing tools rooted in accuracy, transparency and the structured reporting practices followed by national Veterinary Services. AI could complement, rather than replace, existing systems, supporting anomaly detection, prioritising data checks, and ultimately strengthening the reliability and efficiency of global reporting [3]. 

Some key WAHIS processes, such as consistency checks during report verification, remain partly manual. This involves reviewing new submissions to ensure they align with historical patterns and do not contain discrepancies, such as sudden changes in species affected, case counts or geographical distribution. During the heightened avian influenza activity of 2024–2025, outbreak reports submitted through WAHIS nearly doubled. Seasonal spikes in high pathogenicity avian influenza strained WOAH’s capacity, revealing the limits of human-only verification to meet the goal of publishing validated reports within 24 hours of submission [4]. 

The strength of AI lies in its ability to handle structured, repetitive tasks with speed and precision [5]. In WAHIS, AI could help verify reports more quickly, detect inconsistencies and support countries in meeting reporting standards [6]. As WAHIS Project Manager, I see strong potential in pursuing these opportunities. Of course, while WAHIS rests on a robust foundation, any AI-enabled enhancements must be grounded in sound governance, high-quality data and the right expertise to ensure innovation is both effective and responsible.

Considerations Before Implementing AI in WAHIS

1. Data Quality: The Foundation

WAHIS benefits from more than two decades of verified data from Veterinary Authorities worldwide – a rare asset in both public and animal health systems. After all, AI is only as good as the data it draws on. Accuracy, timeliness and contextual detail, covering both quantitative data and control measures, are essential for informed decision-making.  

2. Explainability and Trust

AI can often deliver results that are difficult to interpret. In a system like WAHIS, where data drives public health responses and political decisions, lack of transparency is unacceptable. Any AI-assisted workflow must be explainable: users should understand why a report was flagged – whether for missing data, inconsistencies, or unusual patterns – not merely that it was flagged. Transparency is crucial for building trust and ensuring accountability stays with humans, not machines [7]. 

3. Bias and Equity

Bias is not a theoretical risk. If AI models are trained on datasets that underrepresent certain geographical regions or animal species, they may overlook important patterns or reinforce existing blind spots. In a global platform like WAHIS, this could have serious consequences. AI tools must be continuously audited for fairness across geographic regions and species to ensure that innovation benefits all WOAH Members equally, not just those with advanced animal health surveillance and reporting systems [8]. 

4. AI and Multilateralism

WAHIS is a collective global effort. Countries trust the platform to reflect their data fairly, safeguard their inputs and support national systems. AI should reinforce this trust and collaboration, not disrupt it [9]. Designed inclusively and transparently, AI could improve consistency, reduce manual workloads and foster more equitable engagement with WAHIS [10]. However, its success depends on sustained trust, clear communication and inclusive, responsible governance [11]. 

Piloting AI in WAHIS: A Practical Next Step

As WOAH continues exploring the potential of AI, a logical next step is to launch pilot projects within the WAHIS environment. These pilots would provide a safe, structured way to test AI’s value in supporting core workflows. WOAH has already laid important foundations through broader digital transformation initiatives, including a secure, scalable infrastructure hosted on Microsoft Azure, serving both WAHIS and the Performance of Veterinary Services (PVS) Pathway Information System [12].  

If approved, the pilots would assess AI’s capacity to enhance key WAHIS functions such as report verification, anomaly detection and support for WOAH Members and non-Members in meeting reporting standards [6]. They would not replace existing processes; rather, they would test how AI can complement WOAH’s human expertise in verifying submitted reports, by handling routine checks and surfacing early signals [9]. 

1. Pilot Design Objectives

Pilot projects would be designed to: 

• Test AI performance under real-world WAHIS conditions, including varied reporting formats, incomplete or delayed data submissions, and the need for consistency checks across historical records. 

• Compare AI outcomes with existing manual or semi-manual processes. 

• Evaluate resilience to real-world data gaps (for example, assessing whether AI tools can still produce useful or safe recommendations when information is missing, delayed or inconsistent). 

• Gather ongoing feedback from WAHIS reporting users (WOAH Delegates and national Focal Points), WOAH technical staff and external experts. 

Early involvement of technical staff and Veterinary Authorities will help ensure the AI tools address real needs and improve through iterative refinement. A human-in-the-loop approach is essential to building systems that are both effective and trustworthy [13]. 

2. What the Pilots Aim to Clarify

Before considering broader deployment, these exploratory pilots would help WOAH answer key questions: 

• Can AI tools perform consistently in dynamic, imperfect reporting conditions? 

• What level of accuracy is acceptable when public and animal health decisions are at stake? 

• How can transparency and traceability be ensured in automated recommendations? 

• How will WOAH Members respond to outputs from systems they do not directly control? 

These pilots would assess AI’s role in enhancing report validation, anomaly detection and support to WOAH Members and non-Members, always as a complement to, not a replacement for, current workflows. They would serve as a bridge between today’s system and a future of responsible, AI-assisted surveillance. 

Seizing the Opportunity for Responsible AI

Global animal health challenges are intensifying: more data, more outbreaks, greater urgency. Although WAHIS was launched in 2005, it builds on a century of trusted cooperation between WOAH and national authorities in disease notification [14]. With its robust foundation of verified data, WAHIS is uniquely positioned to lead a new phase of innovation. AI is not a shortcut; it is a tool. Used wisely, it can help WOAH address emerging challenges through pilot-led, expert-informed development. As WAHIS Project Manager, I see a rare opportunity for WOAH to shape the future of AI-assisted animal disease surveillance. 

Main image: ©KTM_2016

DOI: https://doi.org/10.20506/woah.3637

References

[1] World Organisation for Animal Health (WOAH). World Animal Health Information System (WAHIS). Paris (France): WOAH; 2025. Available at: https://wahis.woah.org (accessed on 10 June 2025).

[2] World Organisation for Animal Health (WOAH). Members. Paris (France): WOAH; 2025. Available at: https://www.woah.org/en/who-we-are/members (accessed on 6 August 2025).

[3] The data backbone: building robust AI ecosystems for modern pet care [Internet]. The Digital PawPrint; 2025. Available at: https://pawprint.digital/p/data-backbone-building-ai-ecosystems (accessed on 10 June 2025).

[4] World Organisation for Animal Health (WOAH). Avian influenza. Paris (France): WOAH; 2025. Available at: https://www.woah.org/en/disease/avian-influenza/#ui-id-2 (accessed on 16 July 2025).

[5]  VetCT. Artificial intelligence in veterinary medicine. Cambridge (United Kingdom): VetCT; 2025. 63 p. Available at: https://5345458.fs1.hubspotusercontent-na1.net/hubfs/5345458/AI%20White%20Paper_Final_Web_Foreword.pdf (accessed on 25 June 2025).

[6] World Organisation for Animal Health (WOAH). Codes and Manuals. Paris (France): WOAH; 2025. Available at: https://www.woah.org/en/what-we-do/standards/codes-and-manuals (accessed on 16 July 2025).

[7] European Commission. Ethics guidelines for trustworthy AI. Brussels (Belgium): European Commission; 2019. 41 p. Available at: https://digital-strategy.ec.europa.eu/en/library/ethics-guidelines-trustworthy-ai (accessed on 24 June 2025).

[8] Mehrabi N, Morstatter F, Saxena N, Lerman K, Galstyan A. A survey on bias and fairness in machine learning. ACM Comput. Surv. 2021;54(6):1-35. https://doi.org/10.1145/3457607

[9] Lin S-Y, Beltrán‐Alcrudo D, Awada L, Hamilton‐West C, Lavarello Schettini A, Cáceres P, et al. Analysing WAHIS animal health immediate notifications to understand global reporting trends and measure early warning capacities (2005–2021). Transbound. Emerg. Dis. 2023;2023(1):6666672.  https://doi.org/10.1155/2023/6666672

[10] Gowthami D, Aakash K, Arsath AM, Kumar VS. Smart animal health monitoring system using IoT. Int. Res. J. Educ. Technol. 2023;05(05):247-59. Available at: https://www.irjweb.com/Smart%20Animal%20Health%20Monitoring.pdf (accessed on 10 June 2025).

[11] Hammond A. Balancing AI innovation with responsible governance. Paris (France): World Organisation for Animal Health; 2025. https://doi.org/10.20506/woah.3632

[12] Lasley JN, Alessandrini B, Raza Mirza W, Abdelsattar H. Smart text, stronger systems: large language models for the Performance of Veterinary Services. Paris (France): World Organisation for Animal Health; 2025. https://doi.org/10.20506/woah.3640

[13] Sun JJ. Toward collaborative artificial intelligence development for animal well-being. J. Am. Vet. Med. Assoc. 2025;263(4):528-35. https://doi.org/10.2460/javma.24.10.0650

[14] World Organisation for Animal Health (WOAH). WOAH turns 100: a century of improving animal health and welfare. Paris (France): WOAH; 2025. Available at: https://www.woah.org/en/woah-turns-100-a-century-of-improving-animal-health-and-welfare (accessed on 6 August 2025).