Adversarial Attacks on FinTech AI Models: Threats and Mitigation Techniques
Bibliographic record
Abstract
The fast acceptance of Artificial Intelligence technology by Financial Technology operations produced major changes in fraud prevention systems while simultaneously advancing credit evaluation practices and automated trading methods and customer service operations. FinTech systems became more susceptible to adversarial attacks because the recent technological advancement has introduced malicious input designs for deceiving AI models and altering their results. These malicious assaults cause severe damages to financial security while violating both privacy standards and the public belief in AI as a service provider. The research investigates future trends of adversarial attacks aimed at FinTech AI models by examining attack classifications which involve evasion and poisoning and inference attacks together with the specific vulnerabilities found in financial sector machine learning algorithms. Empirical evidence and real-world case studies demonstrate how harmful modifications in fraud detection methods and trading algorithms as well as credit rating assessment techniques become possible via malicious manipulations. An evaluation of contemporary defence approaches including adversarial training, robust optimization, input preprocessing and model uncertainty measures takes place in the paper. This paper evaluates how explainable AI (XAI) together with regulatory frameworks strengthens both transparency and resilience of systems. The research advances understandings about safeguarding financial services AI while delivering practical steps that FinTech practitioners and policy officials and academic researchers need to create reliable trustable AI environments in FinTech.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".