Journal article · 2026 · Open access

An adaptive differential privacy framework for clinical llms with context-aware noise calibration, hierarchical budgeting, and real-time auditing

Ans D. Alghamdi

Scientific Reports, 16(1) · Published 2 April 2026 · DOI 10.1038/s41598-026-45883-6

Open PDFView at the publisherCite this paper

In brief

“An adaptive differential privacy framework for clinical llms with context-aware noise calibration, hierarchical budgeting, and real-time auditing” is a 2026 journal article by Ans D. Alghamdi, published in Scientific Reports. OpenAlex counts 3 citations as of 19 November 2026, which puts it in the top 4% of works published in 2026 by citations. It is open access under a CC BY-NC-ND licence: the full text is free to read at Scientific Reports.

Citations
3
Most cited in
2026 (3)
Field-weighted impact
3.22
References
38
Authors
1

Abstract

Ensuring that artificial intelligence can support clinical report generation without compromising patient privacy remains a critical challenge in healthcare AI. To address this issue, we propose a novel framework (PrivLLM-Guard) for differentially private large language models (LLMs) designed for real-time confidential medical text generation and summarization. While LLMs have demonstrated substantial potential in automating clinical documentation, the extreme sensitivity of healthcare data demands rigorous, formally grounded privacy guarantees. Our proposed framework addresses this need by combining advanced (ε,δ)-differential privacy techniques with three key innovations: (i) an adaptive noise calibration system that dynamically adjusts Gaussian noise parameters based on input sensitivity and real-time privacy risk scores, (ii) a hierarchical privacy budget allocation mechanism that assigns differentiated protection levels to medical data categories (e.g., patient identifiers, diagnoses, demographics) using Rényi Differential Privacy (RDP) accounting for tight composition bounds during long-sequence generation, and (iii) an integrated real-time privacy auditing module that continuously monitors information leakage probabilities and triggers adaptive mitigation responses. The framework integrates bidirectional transformer encoders with autoregressive decoders, further enhanced by privacy-aware attention mechanisms with calibrated noise injection (Eq. 5) and gradient perturbation via differentially-private SGD with adaptive clipping (Eqs. 3–4). Extensive experiments on three large-scale medical datasets (MIMIC-III with 2.1 M documents, i2b2 with 858 K records, and a proprietary hospital dataset with 1.5 M records) under strict privacy constraints (ε = 0.1, δ = 10⁻⁶) demonstrate BLEU-4 scores of 89.7% for generation and ROUGE-L scores of 92.3% for summarization, representing a 16.8% improvement over the best baseline. The model processes 512-token sequences in real time with an average latency of 245 ms, a throughput of 19.3 requests per second, and memory usage of just 4.2 GB. Formal computational complexity analysis shows O(n²·d + n·κ) inference cost, where n is sequence length, d is model dimension, and κ represents per-token noise calibration overhead. Compared to state-of-the-art privacy-preserving LLMs, By improving the utility–privacy trade-off by 15.8%, mitigating membership inference risk by 65.9%, and reducing computational overhead by 23.4%, the framework establishes a more secure, efficient, and trustworthy foundation for clinical AI deployment.

Cite this paper

Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.

APA 7

Alghamdi, A. D. (2026). An adaptive differential privacy framework for clinical llms with context-aware noise calibration, hierarchical budgeting, and real-time auditing. Scientific Reports, 16(1). https://doi.org/10.1038/s41598-026-45883-6

MLA 9

Alghamdi, Ans D. “An adaptive differential privacy framework for clinical llms with context-aware noise calibration, hierarchical budgeting, and real-time auditing.” Scientific Reports, vol. 16, no. 1, 2026. https://doi.org/10.1038/s41598-026-45883-6.

Chicago

Alghamdi, Ans D. 2026. “An adaptive differential privacy framework for clinical llms with context-aware noise calibration, hierarchical budgeting, and real-time auditing.” Scientific Reports 16 (1). https://doi.org/10.1038/s41598-026-45883-6.

Harvard

Alghamdi, A. D. (2026) ‘An adaptive differential privacy framework for clinical llms with context-aware noise calibration, hierarchical budgeting, and real-time auditing’, Scientific Reports, 16(1). doi:10.1038/s41598-026-45883-6.

IEEE

A. D. Alghamdi, “An adaptive differential privacy framework for clinical llms with context-aware noise calibration, hierarchical budgeting, and real-time auditing,” Scientific Reports, vol. 16, no. 1, 2026, doi: 10.1038/s41598-026-45883-6.

BibTeX

@article{alghamdi2026adaptive,
  title     = {{An adaptive differential privacy framework for clinical llms with context-aware noise calibration, hierarchical budgeting, and real-time auditing}},
  author    = {Alghamdi, Ans D.},
  journal   = {Scientific Reports},
  year      = {2026},
  volume    = {16},
  number    = {1},
  doi       = {10.1038/s41598-026-45883-6},
  url       = {https://doi.org/10.1038/s41598-026-45883-6}
}

Download for a reference manager: BibTeX (.bib) · RIS for Zotero, Mendeley and EndNote (.ris)

Topic and keywords

Physical Sciences › Computer Science › Artificial Intelligence › Privacy-Preserving Technologies in Data

  • differential privacy
  • clinical large language models
  • clinical text generation
  • clinical summarization
  • privacy budget allocation
  • Rényi differential privacy
  • privacy auditing
  • membership inference

Work with this paper

Most-cited papers in Computer Science

All highly cited Computer Science papers

Questions about this paper

How do I cite this paper in APA style?

Alghamdi, A. D. (2026). An adaptive differential privacy framework for clinical llms with context-aware noise calibration, hierarchical budgeting, and real-time auditing. Scientific Reports, 16(1). https://doi.org/10.1038/s41598-026-45883-6

How many times has this paper been cited?

3 times, according to OpenAlex on 19 November 2026. Google Scholar and Scopus count different sets of sources, so their numbers differ.

Is this paper free to read?

Yes. It is open access at Scientific Reports: sign in and open the PDF from this page.

Who wrote this paper?

Ans D. Alghamdi.

What is the DOI of this paper?

10.1038/s41598-026-45883-6, permanently at https://doi.org/10.1038/s41598-026-45883-6. Sign in and open it from this page.

Bibliographic data from OpenAlex (CC0), last checked 19 November 2026. ResearchGuru does not host this paper; once you are signed in, the links open it at the publisher or the repository that does.