Truace
Subscribe
The Index recomputed live from the record

What the evidence says.What the public feels.

The record holds 933 sourced gains and 770 sourced problems, averaging 68 and 66 on the index score. Readers have logged 22 public signals on the Pulse, which is kept apart and never counted as evidence.

933 gains
770 problems
Every sourced claim in the Index, one square each, shaded by the strength of its evidence. High Moderate Emerging

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

1,703 results
Show filters and sorting

Download every matching row, not just this page:Export CSVExport JSON

AI gains · 933

145
HealthStableModerate evidence · 1 source

A point-of-care exhaled breath condensate device combined with physiological parameters and ensemble machine learning identified early-stage lung cancer in a real-world screening cohort with 85.7% accuracy and 100% specificity on held-out test data.

Researchers tested Inflammacheck, a point-of-care device that measures hydrogen peroxide in exhaled breath condensate plus physiological signals, combined with machine learning, in 34 participants from a UK lung health check programme where 83% of cancers were stage I-II. Multivariate analyses separated cancer and control groups, and a voting ensemble achieved 85.7% accuracy and 0.90 ROC-AUC on held-out data.

Impact 30%
69
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
86

Updated Aug 1, 2026 · TRV-2026-0616

73Index score
146
HealthStableModerate evidence · 1 source

Machine learning models using structural MRI, amyloid PET, and demographic features can classify tau positivity in the Braak III/IV region in amyloid-positive cohorts, achieving AUC 0.92 and 85% accuracy on external validation.

By August 2026, researchers had trained machine learning models on ADNI data to predict tau PET positivity from more accessible MRI and amyloid PET features, then tested them on OASIS-3 and SCAN cohorts. Logistic regression reached AUCs of 0.92 in both internal and external validation, with combined external accuracy of 85%.

Impact 30%
69
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
86

Updated Aug 1, 2026 · TRV-2026-0610

73Index score
147
ScienceStableModerate evidence · 1 source

LLM-driven adaptive level modification framework classified players by skill with 97.82% accuracy and generated modified levels that remained traversable at 74.1% full-level and 83.5% chunk-level rates.

On July 28, 2026, Scientific Reports published a framework for adaptive level modification that continuously infers player skill and restructures game content in real time. The system combines reinforcement learning agents and human data to classify skill, then uses a two-stage large language model pipeline to rewrite level chunks, with a physics-constrained verifier to preserve playability.

Impact 30%
69
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
86

Updated Jul 30, 2026 · TRV-2026-0594

73Index score
148
HealthStableModerate evidence · 1 source

A machine learning model using bedside fNIRS functional connectivity during audio movie clips predicted 6-month functional outcome in ICU patients with acute brain injury with 81.3% balanced accuracy, outperforming clinical models.

In a prospective cohort of 33 ICU patients with acute brain injury, investigators tested whether bedside functional near-infrared spectroscopy during passive audio movie listening could support early prognostication. Using functional connectivity features to train a machine learning model, they classified 6-month functional outcome defined by Glasgow Outcome Scale-Extended.

Impact 30%
69
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
85

Updated Jul 26, 2026 · TRV-2026-0572

73Index score

AI problems · 770

145
CrimeStableHigh evidence · 5 sources

Perpetrators obtain children's photos from social media and use AI-based applications to manipulate them into indecent content constituting digital pornography.

By June 2026, researchers documented a cybercrime pattern in Cirebon where perpetrators obtained children's photographs from social media or other digital platforms and manipulated them using Artificial Intelligence-based applications to produce indecent content.

Impact 30%
49
Evidence 25%
100
Scale 20%
35
Confidence 15%
100
Recency 10%
83

Updated Jul 13, 2026 · TRV-2026-0130

70Index score
146
LaborStableHigh evidence · 5 sources

AI-enabled offensive techniques increase cybersecurity risk in Asia-Pacific by using data poisoning, model extraction and AI-driven social engineering, with agentic AI expanding attack surfaces

By July 2026 this peer-reviewed synthesis examined how artificial intelligence reshapes cybersecurity in the Asia-Pacific, focusing on ASEAN. It catalogued offensive methods such as data poisoning and model extraction and defensive uses of machine learning and deep learning for anomaly detection and incident prioritization, alongside technical controls like sandboxing and identity controls.

Impact 30%
49
Evidence 25%
100
Scale 20%
35
Confidence 15%
100
Recency 10%
83

Updated Jul 13, 2026 · TRV-2026-0124

70Index score
147
HealthRisingHigh evidence · 5 sources

AI systems in healthcare have seen limited successful deployment into clinical practice due to intrinsic machine learning limitations and implementation barriers

Researchers reviewed the state of AI in healthcare, noting rapid acceleration of research and demonstrations across medical domains but few cases where techniques have moved into routine clinical use. The article frames translation as the central issue and outlines categories of obstacles that prevent research models from reaching practice.

Impact 30%
49
Evidence 25%
100
Scale 20%
35
Confidence 15%
100
Recency 10%
83

Updated Jul 12, 2026 · TRV-2026-0059

70Index score
148
LaborNewly addedModerate evidence · 1 source

Emotional AI companions risk replacing authentic human intimacy with algorithmic pseudo-intimacy, fostering emotional dependence, sustained loneliness, and displacement of human ties.

A peer-reviewed article in Frontiers in Psychology published September 18, 2025 examines emotional AI systems including Replika, Xiaoice, companion chatbots and therapeutic apps that interpret emotional states and simulate empathic responses, marketed as AI friends for users facing isolation.

Impact 30%
49
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
100

Updated Oct 10, 2026 · TRV-2026-1320

69Index score

Recomputed live from the record · Oct 11, 2026, 6:16 PM