Nos derniers articles

Recent #ai developments, particularly in Natural Language Processing (#nlp) like #gpt3, are widely used. Ensuring safety and trust with increasing NLP use requires robust guidelines. Global AI #regulations are evolving through initiatives like the #euaiact, #unesco recommendations, #us AI Bill of Rights, and others. The EU AI Act's comprehensive regulation sets a potential global benchmark. NLP models are subject to existing rules, such as #gdpr. This paper explores AI regulations, GDPR's application to AI, the EU AI Act's #riskbasedapproach, and NLP's role within these frameworks.

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.

The study proposes a method to assess #demographic #risk within the #solvencyii #regulations, using compact formulas to analyse #insurance portfolio inflows and outflows. It recommends a market-consistent valuation of liabilities for traditional and equity-linked policies. This includes evaluation of the Solvency #capitalrequirement of idiosyncratic and systematic risk, with a formula for the former and an algorithm for the latter.

"#banks take costly actions (such as higher #capitalization, #liquidity holding, and advanced #riskmanagement) to avoid financial distress and #bankruns ... We show that #prudential #regulations have an informational impact: sufficiently tight regulations can eliminate inefficient separating equilibria in banks’ signaling game, thereby changing the information available to creditors and their incentives to run."

This paper addresses the challenges associated with the adoption of #machinelearning (#ml) in #financialinstitutions. While ML models offer high predictive accuracy, their lack of explainability, robustness, and fairness raises concerns about their trustworthiness. Furthermore, proposed #regulations require high-risk #ai systems to meet specific #requirements. To address these gaps, the paper introduces the Key AI Risk Indicators (KAIRI) framework, tailored to the #financialservices industry. The framework maps #regulatoryrequirements from the #euaiact to four measurable principles (Sustainability, Accuracy, Fairness, Explainability). For each principle, a set of statistical metrics is proposed to #measure, #manage, and #mitigate #airisks in #finance.

There are five different common reactions to dealing with, or taming, this #uncertainty in #cyberspace: (1) using #riskmanagement to control uncertainty; (2) recovering from uncertainty through #resilience; (3) mitigating uncertainty through the use of #laws and #regulations; (4) suspending uncertainty by engaging in trust; and (5) ignoring uncertainty through inaction.

This study examines the use of #artificialintelligence (#ai) and #bigdata data analytics by #insurers in #belgium for segmentation purposes to determine #claims#probability for prospective policyholders. The implementation of AI and big data analytics can benefit insurers by increasing the accuracy of #riskassessment. However, pervasive segmentation can have negative implications and potentially harm policyholders if their risk is incorrectly calculated. Existing restrictions in #insurance#regulations fall short of protecting policyholders from inaccuracies in risk assessments, potentially resulting in incorrect #premiums or conditions.

The study highlights that while modern states have developed concrete strategies to respond to potential threats, the resemblance of these strategies to one another could create unexpected challenges. The dynamic nature of the internet and the multitude of actors and sources of risk could put conventional wisdom to the test at a stage where the scope for response is limited. This highlights the need for states to continually adapt their strategies to address emerging risks and avoid relying solely on common knowledge or uniform thinking.

"Shocks to disaster costs seem to decrease all type of emissions significantly and also increase renewable energy use significantly. The occurrence of natural disasters increases the political disagreement among U.S. politicians, as well as, the climate policy uncertainty, highlighting the need for efficient policymaking and regulations. "

"The European Artificial Intelligence Board (EAIB) would be established as a new enforcement authority at the Union level. National supervisors will flank EAIB at the Member State level. Fines of up to '6% of global turnover, or 30 million euros for individual corporations' can be imposed."