|
所在平台: Udemy |
课程主页: https://www.udemy.com/course/fraud-with-gen-ai-defense/
课程评论：没有评论
课程名称：抵御基于生成式人工智能的欺诈 课程概述： 欺诈（包括支付欺诈和保险欺诈）是组织面临的最重大问题之一。随着生成式人工智能的进步，这一问题变得更加复杂。在当今世界，组织和个人不仅必须抵御欺诈尝试，还必须对抗利用生成式人工智能的欺诈行为，这通常意味着更快、更大规模和更复杂的攻击。本课程将教你如何保护自己免受基于生成式人工智能的欺诈。 课程内容： - 学习生成式人工智能的基础知识，包括常见模型和模型家族，生成内容的特征，以及它可能因疏忽或恶意行为而被滥用的方式（如偏见、误信息、身份盗用等）。 - 学习欺诈的基本概念和主要类型（如身份盗窃、支付欺诈、投资欺诈、保险欺诈、账户接管），以及促进这些欺诈的主要因素（如技术差距、人为错误、流程弱点、数据泄露）。 - 理解生成式人工智能加速欺诈的方式（如大规模自动化、提高真实性、模式规避、合成身份创建等）以及其对主要欺诈方法的影响（如文件伪造、交易操控、合成身份欺诈、索赔欺诈等）。 - 了解主要的生成内容类型及其在欺诈攻击中的使用（文本、图像、音频和视频），包括每种类型利用的具体方法、训练模型所需的数据以及如何检测每种类型的欺诈。 - 生成文本：学习支持生成文本的模型（如大型语言模型）、传播渠道（如电子邮件、短信、聊天）、训练所需数据及检测机制（如行为分析、身份验证和文本验证）。 - 生成图像：学习支持生成图像的模型（如生成对抗网络或扩散模型）、传播渠道（如欺诈性文档和电子邮件附件）、训练所需数据，以及检测机制（如水印、行为检测或多因素身份验证）。 - 生成音频：学习支持生成音频的模型（如生成对抗网络或文字转语音）、传播渠道（如语音消息软件或电话）、训练所需数据，以及检测机制（如多因素身份验证因素、回拨或员工培训）。 - 生成视频：学习支持生成视频的模型（如用于视频的生成对抗网络或深度学习模型）、传播渠道（如通讯工具或提交门户）、训练所需数据，以及检测机制（如验证通讯、反深度伪造软件、多因素身份验证）。 本课程将为你提供抵御基于生成式人工智能的欺诈的系统知识和技能。
BEATING FRAUDFraud, including payment fraud and insurance fraud, is one of the biggest problems for organizations.And new advances in terms of generative AI have only made this worse.In the world of today, organizations and individuals must be able to not only resist fraud attempts, but resist when they leverage generative AI - which, in many cases, means faster, larger-scale and more sophisticated attacks.This course will teach you how to protect against fraud that leverages generative AI.LET ME TELL YOU.EVERYTHING.Some people - including me - love to know what they're getting in a package.And by this, I mean, EVERYTHING that is in the package.So, here is a list of everything that this course covers:You'll learn the basics of generative AI and what it can do, including common models and families of models, the characteristics of generative content, and how it can be misused due to negligence or active malevolence (including biases, misinformation, impersonation and more);You'll learn the basics of fraud and its main types (identity theft, payment fraud, investment fraud, insurance fraud, account takeover), and the main factors enabling it (technological gaps, human error, process weaknesses, data breaches);You'll learn how fraud is accelerated by generative AI (mass automation, increased authenticity, pattern evasion, synthetic identity creation, etc) and its effect on the major approaches (document forgery, transaction manipulation, synthetic identity fraud, claims fraud, etc);You'll learn about an overview of the major generative content types used in fraud attacks (text, image, audio and video), including the specific approaches that each leverage, the model training requirements and data required for attackers to train such models, and how each type can be detected;You'll learn about generative text in fraud, including the models that allow it such as LLMs, the distribution channels such as email, SMS, chats, the data required to train such models, and detection mechanisms such as behavioral analysis, authentication and text validations;You'll learn about generative image in fraud, including the models that allow it such as GANs or diffusion models, the distribution channels such as deceptive documents and images in email attachments or submission portals, the data required to train such models, and detection mechanisms such as watermarks, behavioral detection or MFA;You'll learn about generative audio in fraud, including the models that allow it such as GANs or TTS, the distribution channels such as voice messaging software or for calls, the data required to train such models, and detection mechanisms such as MFA factors, callbacks, or training employees;You'll learn about generative video in fraud, including the models that allow it such as GANs for video or deep learning models, the distribution channels such as communication tools or submission portals, the data required to train such models, and detection mechanisms such as verifying communications, anti-deepfake software, MFA;
