Artificial Intelligence (AI) Trust Framework and Maturity Model: Applying an Entropy Lens to Improve Security, Privacy, and Ethical AI.

Recent advancements in artificial intelligence (AI) technology have raised concerns about the ethical, moral, and legal safeguards. There is a pressing need to improve metrics for assessing security and privacy of AI systems and to manage AI technology in a more ethical manner. To address these challenges, an AI Trust Framework and Maturity Model is proposed to enhance trust in the design and management of AI systems. Trust in AI involves an agreed-upon understanding between humans and machines about system performance. The framework utilizes an "entropy lens" to root the study in information theory and enhance transparency and trust in "black box" AI systems, which lack ethical guardrails. High entropy in AI systems can decrease human trust, particularly in uncertain and competitive environments. The research draws inspiration from entropy studies to improve trust and performance in autonomous human-machine teams and systems, including interconnected elements in hierarchical systems. Applying this lens to improve trust in AI also highlights new opportunities to optimize performance in teams. Two use cases are described to validate the AI framework's ability to measure trust in the design and management of AI systems.

Opportunistic viral surveillance confirms the ongoing disease threat grey squirrels pose to sympatric red squirrel populations in the UK.

BACKGROUND: Red Squirrels United was a UK landscape-scale grey squirrel management programme undertaken between 2016 and 2020. METHODS: A total of 11034 grey squirrels were removed by culling, with 1506 necropsied and 1405 suitable for adenovirus (AdV) or squirrelpox virus (SQPV) quantitative PCR (qPCR) analysis. Spleen, lip or hair were extracted, and DNA was isolated, with samples tested in duplicate by qPCR. RESULTS: Of 1378 tissue samples, 43% were positive for AdV and 10% for SQPV. Of 1031 hair samples, 11% were positive for AdV and 10% for SQPV. Overall, 762 of 1405 (54%) animals were positive for one or both viruses. LIMITATIONS: Ad hoc sampling was undertaken from limited geographical areas but provided the only dataset from that period, instead of extrapolating from historical data. CONCLUSIONS: The grey squirrel is an asymptomatic reservoir host for AdV and SQPV. Interspecific infection transmission potential is demonstrated. Grey squirrel management by culling is essential for mainland red squirrel viability until other suitable management tools are available.

CD24 is a genetic modifier for risk and progression of multiple sclerosis.

Multiple sclerosis (MS) is a chronic neurological disease of unknown etiology, but a genetic basis for the disease is undisputed. We have reported that CD24 is required for the pathogenicity of autoreactive T cells in experimental autoimmune encephalomyelitis, the mouse model of MS. Here we investigate the contribution of CD24 to MS by studying single-nucleotide polymorphism in the ORF among 242 MS patients and 207 population controls. This single-nucleotide polymorphism results in replacement of alanine (CD24a) with valine (CD24v) in the mature protein. We found that the CD24v/v renders a >2-fold increase in the relative risk of MS in the general population (P = 0.023). Among familial MS, the CD24v allele is preferentially transmitted into affected individuals (P = 0.017). Furthermore, 50% of CD24v/v patients with expanded disability status scale 6.0 reached the milestone in 5 years, whereas the CD24a/v (P = 0.00037) and CD24a/a (P = 0.0016) patients did so in 16 and 13 years, respectively. Moreover, our data suggest that the CD24v/v patients expressed higher levels of CD24 on peripheral blood T cells than did the CD24a/a patients. Transfection with CD24a and CD24v cDNA demonstrated that the CD24v allele can be expressed at higher efficiency than the CD24a alleles. Thus, CD24 polymorphism is a genetic modifier for susceptibility and progression of MS in the central Ohio cohort that we studied, perhaps by affecting the efficiency of CD24 expression on the cell surface.

Concordant down-regulation of proto-oncogene PML and major histocompatibility antigen HLA class I expression in high-grade prostate cancer.

Recognition of tumor cells by cytolytic T lymphocytes depends on cell surface MHC class I expression. As a mechanism to evade T cell recognition, many malignant cancer cells, including those of prostate cancer, down-regulate MHC class I. For the majority of human cancers, the molecular mechanism of MHC class I down regulation is unclear, although it is well established that MHC class I down-regulation is often associated with the down-regulation of multiple genes devoted to antigen presentation. Since the promyelocytic leukemia (PML) proto-oncogene controls multiple antigen-presentation genes in some murine cancer cells, we analyzed the expression of proto-oncogene PML and MHC class I in high-grade prostate cancer. We found that 30 of 37 (81%) prostate adenocarcinoma cases with a Gleason grade of 7-8 had more than 50% down-regulation of HLA class I expression. Among these, 22 cases (73.3%) had no detectable PML protein, while 4 cases (13.3%) showed partial PML down-regulation. In contrast, all 7 cases of prostate cancer with high expression of cell surface HLA class I had high levels of PML expression. Concordant down-regulation of HLA and PML was observed in different histological patterns of prostate adenocarcinoma. These results suggest that in high-grade prostate cancer, malfunction of proto-oncogene PML is a major factor in the down-regulation of cell surface HLA class I molecules, the target molecules essential for the direct recognition of cancer cells by cytolytic T lymphocytes.