GENXAI A RESPONSIBLE AI GOVERNANCE FRAMEWORK FOR GENERATIVE MODELS IN REGULATED INDUSTRIES

GENXAI: система ответственного управления ИИ для генеративных моделей в регулируемых отраслях
Elangovan Sivalingam
2026-02-09

data lineagegenerative modelshuman-in-the-loop validationregulated industriesresponsible AI governance
The rapid adoption of generative AI in regulated industries as a result of the media coverage it receives has also shown there are very large holes and gaps in our systems and processes for governing, policing and auditing its use.Classic generative models are a well-known class of machine learning models that, however, still have two strange properties: they base the output behavior on probabilities and it is generally hard to understand the generated content itself.We present GENXAI, a responsible AI governance lifeycle framework for generative model development, deployment and operation in regulated industries.GENXAI leverages a life-cycle governance model to embed compliance, ethical risk reduction and explainability into the generative systems workflows.It does this by marrying the following: Data lineage driven governance, Model transparency and human in the loop validation, Iterative monitoring and audit based generation of auditable documents along to keep constantly compliant.One of the most relevant contributions of this work is its centralized governance model, which maps the abstract regulatory and ethical principles into the technical and organizational mandates.GENXAI is applied with analytical comparisons as well as domain-specific use cases on healthcare, the financial services and legal systems. GENXAI A Responsible AI Governance Framework for Generative Models in Regulated Industries
1
GENXAI combines data-lineage-driven governance, model transparency, human-in-the-loop validation, iterative monitoring, and audit-based generation of compliance documents.
2
GENXAI introduces a responsible AI governance lifecycle framework for developing, deploying, and operating generative models in regulated industries.
3
Its centralized governance model translates abstract regulatory and ethical principles into concrete technical and organizational requirements.
4
The framework embeds compliance, ethical-risk reduction, and explainability directly into generative AI workflows through lifecycle governance.
5
The framework is evaluated through analytical comparisons and domain-specific use cases in healthcare, financial services, and legal systems.

Generative AI models and their development, deployment, and operation in regulated industries

Lifecycle governance of generative AI, focusing on compliance, ethical risk reduction, explainability, transparency, human validation, monitoring, and auditability

Publication Details
Publication Date
2026-02-09
Journal
Publisher
ISSN
Cited by
1
Access Type
Author Information
Authors
Elangovan Sivalingam
Explore further
Open the scid.ai AI chat with a ready-made request: it will find papers on a similar topic and help build a literature review.
Find similar papers in the chat →
Make a presentation
100%