TruaceTracing the truth around AITuesday, September 15, 2026
The Index

What the evidence says.What the public feels.

Ranks distinct AI gain and problem claims from the published record. Scores reward impact, independent source strength, scale, confidence, and recency.

1,419 results
Show filters and sorting

AI gains · 788

68
GainHealth· Stable· Evidence: Moderate (1 source)

AI systems are being adopted in oncology practice to support cancer detection, risk stratification, treatment planning, and clinical documentation workflows.

As of the August 2026 commentary, AI was increasingly integrated into oncology for detection, risk stratification, treatment planning, and documentation. The authors reviewed evidence that these systems can reproduce or amplify disparities and examined technical sources of bias and competing statistical definitions of fairness.

Impact 30%49
Evidence 25%95
Scale 20%35
Confidence 15%87
Recency 10%95

Updated Aug 18, 2026 · TRV-2026-0829

68
GainHealth· Stable· Evidence: Moderate (1 source)

Large language models can restructure prostate MRI reports and extract discrete variables to support supervised summaries and patient-facing explanations.

Published August 17 2026 in Abdominal Radiology, this Perspective examines large language models applied to prostate MRI reporting, a task where laterality, sector, size, PI-RADS, and staging language directly affect biopsy and treatment decisions and where patients often see reports via portals before clinician discussion.

Impact 30%49
Evidence 25%95
Scale 20%35
Confidence 15%87
Recency 10%95

Updated Aug 18, 2026 · TRV-2026-0828

68
GainHealth· Stable· Evidence: Moderate (1 source)

Deep-learning based AIIR reconstruction improved CT image quality and diagnostic accuracy for gastric cancer, increasing tumor conspicuity and raising AUC for detecting serosal invasion compared to hybrid iterative reconstruction.

Researchers prospectively tested a deep-learning based artificial intelligence iterative reconstruction algorithm against conventional hybrid iterative reconstruction in 132 gastric cancer patients undergoing preoperative abdominal CT before surgery or staging laparoscopy. By August 2026, they reported higher Likert scores for tumor margin and enhancement, higher contrast-to-noise ratios in arterial and portal venous phases, and higher AUC for detecting serosal invasion with AIIR.

Impact 30%49
Evidence 25%95
Scale 20%35
Confidence 15%87
Recency 10%95

Updated Aug 18, 2026 · TRV-2026-0827

68
GainHealth· Stable· Evidence: Moderate (1 source)

An integrated framework using WBAF preprocessing, MResU-Net segmentation, IPHOG feature extraction and IShuffleNet-PCNN classification achieved 0.933 accuracy and 0.991 NPV for spinal cord injury-related fracture classification from CT images.

Researchers developed a four-stage deep learning framework for CT-based spinal cord injury fracture assessment, combining Weighted Balanced Anisotropic Filtering for denoising, Modified Residual U-Net for spinal cord segmentation, Improved Pyramid Histogram of Oriented Gradients for feature extraction, and a new IShuffleNet-Parallel CNN classifier with Group Normalization and Adaptive Swish-Mish activation.

Impact 30%49
Evidence 25%95
Scale 20%35
Confidence 15%87
Recency 10%95

Updated Aug 18, 2026 · TRV-2026-0826

AI problems · 631

67
ProblemEducation· Stable· Evidence: Moderate (1 source)

Generative AI chatbots in higher education raise concerns around academic integrity, labour displacement, and embedded biases that require a critical approach.

The peer-reviewed commentary from April 2024 analyzes the rapid uptake of generative AI chatbots like ChatGPT in universities, noting both potential to enhance teaching, research, administration and student support and simultaneous pressures from hype and commercial edtech actors.

Impact 30%49
Evidence 25%95
Scale 20%35
Confidence 15%87
Recency 10%89

Updated Jul 20, 2026 · TRV-2026-0438

67
ProblemBusiness· Stable· Evidence: Moderate (1 source)

Existing HRM frameworks fall short of capturing the novel attributes, complexities and impacts of generative AI on workforce dynamics and organizational operations, introducing unique challenges for human resources.

On 2024-04-11 this peer-reviewed paper proposed a strategic human resource management framework for integrating generative artificial intelligence in business, arguing existing HRM frameworks fall short and outlining components including alignment with business objectives, seizing opportunities, resource assessment and orchestration, re-institutionalization and continuous learning.

Impact 30%49
Evidence 25%95
Scale 20%35
Confidence 15%87
Recency 10%89

Updated Jul 20, 2026 · TRV-2026-0437

67
ProblemHealth· Stable· Evidence: Moderate (1 source)

Poorly performing predictive AI models are misleading and may lead to wrong clinical decisions that can be detrimental to patients and increase financial costs, with classification measures being improper at clinically relevant thresholds.

Published in December 2025, this Viewpoint in The Lancet Digital Health evaluates how to validate predictive AI models that estimate binary outcome probabilities for clinical use. The authors reviewed 32 performance measures across five domains and examined whether each measure is proper and whether it accounts for misclassification costs, illustrating findings with the ADNEX model for ovarian tumour malignancy.

Impact 30%49
Evidence 25%95
Scale 20%35
Confidence 15%87
Recency 10%89

Updated Jul 20, 2026 · TRV-2026-0434

67
ProblemLabor· Stable· Evidence: Moderate (1 source)

Gen-AI in creative media and arts industries raises concerns that human creative labour could disappear.

Published May 22, 2024, this peer-reviewed article analyzes generative AI's transformative role in creative media and arts industries through the lens of the 2023 Writers' and Actors' strikes. It critiques the prevailing 'replacing tasks' narrative and applies a meaningful work framework to argue for coexistence that amplifies rather than supplements human creativity, using Jurassic Park's VFX transition as a historical parallel.

Impact 30%49
Evidence 25%95
Scale 20%35
Confidence 15%87
Recency 10%89

Updated Jul 20, 2026 · TRV-2026-0432

Recomputed live from the record · Sep 15, 2026, 5:34 PM