0 cumulative citations
View corpus contextCombining AI with blockchain improves logistics risk coverage and creates auditability: integration catches more risks than AI-only models while blockchain adds tamper detection and end-to-end traceability that neither technology delivers alone.
Assessing the Impact of Blockchain and Artificial Intelligence on Supply Chain Risk Mitigation
openalex
quasi_experimental
medium evidence
7/10 relevance
Summary only
summary available; pdf_status=paywall
Source
PDF
On Olist logistics data, an integrated AI-plus-blockchain system detects more operational risks than AI alone while blockchain contributes tamper-evident traceability and governance capabilities that AI lacks by itself.
Citation observations
Cumulative provider counts captured on specific dates; providers are never combined.
The increasing adoption of artificial intelligence (AI) in supply chain operations is transforming how organizations detect, govern, and respond to risk, raising important questions regarding digital trust, accountability, and transparency.AI models enable predictive risk assessment using large-scale logistics data.However, their outputs are often difficult to audit or independently verify.In contrast, blockchain technology provides immutable and tamper-evident records but lacks predictive capabilities.This study empirically evaluates an integrated AI-Blockchain environment using a real-world e-commerce logistics dataset (Olist Brazilian dataset, September 2016-October 2018).The analysis compares AI-only, blockchain-only, and integrated configurations across both predictive and governance dimensions.Results indicate that the integrated system improves risk detection coverage compared with AI alone, while blockchain adds governance capabilities including tamper detection and end-to-end event traceability.The findings demonstrate that AI and blockchain address different operational failure modes, and their integration creates governance capabilities that neither technology can provide independently.
Summary
Main Finding
Integrating AI predictive models with a blockchain-backed event ledger improves operational risk management in e-commerce logistics: the combined system raises risk-detection coverage relative to AI alone while adding governance features (tamper detection and end-to-end traceability) that blockchain alone cannot provide.
Key Points
- AI strengths: scalable predictive risk assessment using large-scale logistics data; identifies probable failures before they occur.
- AI weaknesses: model outputs are often hard to audit, verify independently, or trace back to raw events.
- Blockchain strengths: immutable, tamper-evident records that enable post-hoc verification and end-to-end event traceability.
- Blockchain weaknesses: no intrinsic predictive capability; by itself does not detect future risks.
- Integrated configuration: couples AI predictions with a blockchain event ledger so that predictive flags, underlying inputs, and relevant events are recorded immutably.
- Empirical result: the integrated system expanded risk-detection coverage relative to AI-only, and provided governance functions (tamper detection, traceability) absent from AI-only or blockchain-only setups.
- Conceptual conclusion: AI and blockchain mitigate different operational failure modes; together they create combined governance and predictive capabilities neither can achieve alone.
Data & Methods
- Dataset: Olist Brazilian e-commerce logistics dataset, covering September 2016–October 2018.
- Experimental configurations compared:
- AI-only: predictive models run on the logistics data, outputs used for risk detection.
- Blockchain-only: events recorded immutably but no predictive layer.
- Integrated AI+Blockchain: AI predictions and associated provenance/events written to the blockchain for immutable linkage.
- Evaluation dimensions:
- Predictive performance (risk-detection coverage and related metrics).
- Governance functionality (tamper detection, end-to-end traceability, auditability, independent verification).
- Findings were based on empirical comparisons across these configurations using real-world operational events; blockchain served as an evidence layer rather than as a substitute for predictive analytics.
- Limitations to note: single-country e-commerce dataset (Brazil) from 2016–2018—generalizability to other sectors, countries, or more recent operational contexts may be limited; practical deployment involves trade-offs around on-chain data vs. off-chain pointers, privacy, throughput, and integration cost.
Implications for AI Economics
- Complementary technologies: AI delivers value through improved operational efficiency and early risk detection; blockchain adds economic value via stronger digital trust, lowering verification costs and enabling credible audit trails.
- Governance and accountability: Immutable recordkeeping can reduce information asymmetries between supply-chain participants and external auditors/regulators, potentially lowering monitoring costs and moral-hazard risks.
- Market structure and contracting: Integrated solutions support more performance-based contracting (pay-for-performance, automated penalties/escrow) because evidence and model outputs are verifiable and tamper-evident.
- Investment trade-offs: Firms should weigh gains in risk coverage and trust against integration costs (engineering, latency, blockchain transaction costs) and privacy-compliance burdens (what to put on-chain).
- Policy and standardization: Regulators and industry consortia may need standards for on-chain data formats, provenance, and model-audit protocols to realize scalable interoperability and reduce verification friction.
- Future research directions: quantifying costs vs. benefits in different industries, designing privacy-preserving on-chain architectures, and formalizing incentive mechanisms that align model operators, data providers, and auditors.
Assessment
Paper Typequasi_experimental
Evidence Strengthmedium — The paper provides empirical, quantitative comparison across AI-only, blockchain-only, and integrated configurations using a real-world logistics dataset, which gives credible within-dataset evidence that integration improves risk detection coverage and governance metrics; however, it is limited to a single historical dataset, relies on simulated integration rather than field deployment, and lacks causal identification methods, external validation, or robustness checks that would support stronger causal claims or broader generalization.
Methods Rigormedium — The study uses real logistics event data and compares measurable predictive and governance outcomes across clearly defined system configurations, but the methods appear to omit randomized or quasi-experimental identification, deployment/operational tests (e.g., adversarial/tampering trials in live systems), sensitivity analyses, and cross-dataset validation that would increase confidence in results and rule out dataset-specific artifacts.
SampleOlist Brazilian e-commerce logistics dataset covering September 2016–October 2018 (marketplace order and shipment event records, timestamps, tracking/status events and associated attributes from a single e-commerce platform/marketplace); used to train/evaluate predictive models and to simulate blockchain event-recording and tamper scenarios.
Themesgovernance adoption
Generalizabilitysingle_country_data (Brazil) — may not generalize to other regulatory/market contexts, single_marketplace/platform (Olist) — results may be platform-specific, historical_period (2016–2018) — logistics processes and AI models have evolved since, simulated blockchain/AI integration rather than live deployment — operational, organizational, and user-behavioral effects unobserved, focus on logistics risk types present in dataset — other failure modes or sectors not evaluated, no evidence on cost, latency, scalability, or runtime performance at production scale, governance outcomes measured technically (tamper-detection, traceability) but not tested against legal/regulatory or organizational response processes
Claims (8)
| Claim | Direction | Outcome | Confidence & Evidence | Details |
|---|---|---|---|---|
| AI models enable predictive risk assessment using large-scale logistics data. Decision Quality | positive | predictive risk assessment capability |
Reading fidelity
high
Study strength
medium
|
not reported
|
| AI model outputs are often difficult to audit or independently verify. Ai Safety And Ethics | negative | auditability / verifiability of AI outputs |
Reading fidelity
high
Study strength
medium
|
not reported
|
| Blockchain provides immutable and tamper-evident records. Governance And Regulation | positive | record immutability / tamper detection |
Reading fidelity
high
Study strength
medium
|
not reported
|
| Blockchain lacks predictive capabilities. Decision Quality | negative | predictive capability |
Reading fidelity
high
Study strength
medium
|
not reported
|
| The integrated AI-Blockchain system improves risk detection coverage compared with AI alone. Decision Quality | positive | risk detection coverage |
Reading fidelity
high
Study strength
medium
|
not reported
|
| Blockchain adds governance capabilities including tamper detection and end-to-end event traceability. Governance And Regulation | positive | tamper detection and event traceability |
Reading fidelity
high
Study strength
medium
|
not reported
|
| AI and blockchain address different operational failure modes, and their integration creates governance capabilities that neither technology can provide independently. Governance And Regulation | positive | coverage of operational failure modes and governance capability creation |
Reading fidelity
high
Study strength
medium
|
not reported
|
| This study empirically evaluates an integrated AI-Blockchain environment using a real-world e-commerce logistics dataset (Olist Brazilian dataset, September 2016–October 2018). Other | positive | empirical evaluation / dataset usage |
Reading fidelity
high
Study strength
high
|
not reported
|