Lessons from GDPR for AI Policymaking

Date : Tags : , , , , , ,
The introduction of #ai #chatgpt has stirred discussions about AI regulation. The controversy over classifying systems like ChatGPT as "high-risk" AI under #euaiact has sparked concerns. This paper explores how Large Language Models (#llms) such as ChatGPT are shaping AI policy debates and delves into potential lessons from the #gdpr for effective regulation.

Regulation of (Generative) AI Requires Continuous Oversight

Date : Tags : , , , , , ,
The submission suggests strategies for regulating #ai in #australia, including examining the rate of take-up of #automated #decisionmaking systems, and regulating specific applications of underlying AI technologies. It also suggests altering the definition of AI, creating a set of guiding principles, and adopting a #risk-based approach to #regulation.

Incident‑Specific Cyber Insurance

"In the current market practice, many #cyberinsurance products offer a coverage bundle for losses arising from various types of incidents, such as #databreaches and #ransomwareattacks, and the coverage for each incident type comes with a separate limit and deductible. Although this gives prospective cyber insurance buyers more flexibility in customizing the coverage and better manages the #risk exposures of sellers, it complicates the decision-making process in determining the optimal amount of risks to retain and transfer for both parties. This paper aims to build an economic foundation for these incident-specific cyber insurance products with a focus on how incident-specific indemnities should be designed for achieving #pareto optimality for both the #insurance seller and buyer. Real data on #cyberincidents is used to illustrate the feasibility of this approach. Several implementation improvement methods for practicality are also discussed."

Climate Risk and Bank Capital Structure

Date : Tags :
"This paper lends support to the importance of the #climatechange-related #risks into #prudential #supervision to protect the #financialsystems #resilience and contributes to the debate on #climate-related #capitalrequirements."

FinPT: Financial Risk Prediction with Profile Tuning on Pretrained Foundation Models

Date : Tags : , , , , , ,
#financialrisk #prediction is vital but hindered by outdated algorithms and the absence of comprehensive benchmarks. Addressing this, FinPT uses large pretrained models and Profile Tuning for #risk prediction, while FinBench provides datasets on #default, #fraud, and #churn. FinPT inserts tabular data into templates, generates customer profiles using #languagemodels, and fine-tunes models for predictions, demonstrated effectively through experiments on FinBench, enhancing understanding of language models in financial risk.

Bayesian and Classical Approaches to Structural Estimation of Risk Attitudes

Date : Tags : , , , ,
This study examines interpersonal heterogeneity in #risk attitudes in #decisionmaking experiments. The use of #bayesian and classical methods for estimating the hierarchical model has sparked debate. Both approaches use the population distribution of risk attitudes to identify individual-specific risk attitudes. Comparing existing experimental data, both methods yield similar conclusions about risk attitudes.

Acceptable Risks in Europe’s Proposed AI Act

This paper critically assesses the proposed #euaiact regarding #riskmanagement and acceptability of #highrisk #ai systems. The Act aims to promote trustworthy AI with proportionate #regulations but its criteria, "as far as possible" (AFAP) and "state of the art," are deemed unworkable and lacking in proportionality and trustworthiness. The Parliament's proposed amendments, introducing "reasonableness" and cost-benefit analysis, are argued to be more balanced and workable.

Fraud Detection by Using Deep Learning in Mining the Information Technology for AI and BI

Date : Tags : , , , ,
The article discusses the use of #deeplearning and #datamining in business intelligence protocols to optimize data-driven decision-making and improve efficiency. The authors focus on the use of Graph Neural Network and Autoencoders Models to process large amounts of data and model #fraud behaviors. They suggest that deep learning can be used to control #moneylaundering in financial institutions and improve visibility and transparency in businesses.

Machine Learning for Automating Monitoring, Review and Testing at Financial Institutions

#financialinstitutions are increasingly using #machinelearningalgorithms for credit risk mgmt., #fraudprevention, and #aml. This paper presents robust evidence of using logistic regression, linear discriminant analysis, and neural networks for accurately predicting and classifying financial transactions for Volcker Rule #compliance. It provides a scalable minimum viable product to automate #controls testing.