GRA · Fintech Nexus
空间金融资本与危害、风险暴露和转型的交汇之处
金融身份
Every financial relationship starts with identifying the customer: proving who they are, checking them against sanctions and risk lists, and authenticating them each time they return. Digital onboarding has widened access and cut cost, but it has also industrialised identity fraud, from synthetic identities to deepfake-assisted account opening. Checks that are too rigid exclude people who lack documents. This page sets out how identity risk develops in finance, how it is measured, and who sets the rules.
财务恒等式:风险如何形成。
客户开户、欺诈与排斥。下方档案阐述风险如何发展、衡量什么、数据来自何处、如何建模以及由谁决定。
Synthetic identities age into credit
欺诈者将真实信息与虚构信息组合成可通过开户审核的身份,经过数月积累良好记录,然后提取信贷并消失。没有真实受害者投诉,因此损失显现较晚,且常被记为违约。
深度伪造攻击远程身份验证
人脸比对和活体检测面临注入式视频、生成的文件和克隆的语音。攻击检测的每一次改进都会遇到新的工具,因此,除非持续测试,身份核验的保证水平会逐渐下降。
Mule accounts carry proceeds
诈骗和欺诈所得资金经由被招募或被冒用身份者开设的账户流转。开户审核薄弱、监测迟缓,使骡账网络得以在冻结生效之前跨机构转移资金。
刚性检查会排除
Document requirements that low-income, migrant, displaced or rural people cannot meet push them out of formal finance. Tiered, risk-based due diligence and trusted digital ID can widen access without lowering integrity.
金融身份对风险暴露、损失和资本意味着什么。
风险官、监管者或财政部会要求的风险暴露、损失、情景和资本视图,以及它们据以报告的框架。结构遵循监管和精算实务。数值为说明性:绝非价格、评级或建议。
传导渠道
How a shock in financial identity reaches credit, market, underwriting, liquidity and operational risk, and feeds back.
衡量什么,以及衡量得如何。
每个指标都注明其单位、频率、背后的来源类型以及数值将附带的证据标签。此处不显示数值:数值由计划依据其自身权限提供。
| 指标 | 单位 | 节奏 | 来源类型 | 标签 |
|---|---|---|---|---|
| Identity fraud at onboarding | 每千份申请中的尝试次数和确认案件数 | 每月 | Firm fraud records | 观测 |
| 账户接管事件 | cases per million active accounts | 每季度 | Firm and industry fraud reporting | 已报告 |
| 呈现攻击检测错误率 | 攻击与真实样本分类错误率 | 每次评估 | Independent biometric testing | 观测 |
| Remote onboarding completion and abandonment | percent of started applications | 每月 | 企业记录 | 观测 |
| 因缺乏证件而未开立账户的成年人 | 没有账户的成年人百分比 | 每轮调查 | Global Findex | 估算 |
| 拥有基础身份证明的人口 | 占人口百分比 | 定期 | ID4D 数据集和国家登记册 | 估算 |
| 与身份欺诈相关的可疑交易报告 | 每年计数 | 年度 | 金融情报机构报告 | 已发布统计 |
证据来自何处。
具名公共数据集与机构记录,各自注明许可类别。依协议获得的来源仅按其条款使用,绝不再次发布。
| 来源 | 它提供什么 | 许可类别 |
|---|---|---|
| 世界银行全球金融包容性数据库(World Bank Global Findex Database) | Account ownership and barriers to access, including lack of documentation, by country and group | Free and open, attribution (CC BY) |
| 世界银行 ID4D 全球数据集 | Coverage of foundational identification systems by country | 免费开放,须署名 |
| NIST 人脸识别和呈现攻击评估 | 生物识别算法的独立准确性和攻击检测结果 | 公有领域 |
| FBI Internet Crime Complaint Center reports | Reported identity theft, impersonation and fraud losses | 公有领域 |
| Financial intelligence unit annual reports | 可疑交易报告的趋势和类型 | Public records, terms per unit |
| FATF guidance on digital identity and mutual evaluation reports | Assessments of customer due diligence regimes and guidance on reliance on digital ID | 免费,条款依 FATF |
由谁决定,以及证据服务于什么。
Fintech Nexus 负责准备证据。以下决定由所列机构依其自身规则作出,平台绝不代为决定。
| 谁来决定 | 问题 | 本页面服务的证据 |
|---|---|---|
| 金融情报机构和反洗钱监管机构 | 哪些客户尽职调查方法符合规则,以及在哪些情形下允许采取简化措施 | Assurance-level mapping, typology notes, inclusion evidence |
| National identity authorities | 金融机构能否以及如何依赖国家数字身份办理客户开户 | 保证映射、覆盖缺口、欺诈与错误记录 |
| Bank and fintech boards | 应资助哪些身份证明、活体检测和监测控制措施,以及可接受多大的客户摩擦 | 开户环节的欺诈、攻击检测结果、放弃率 |
| 标准机构(FATF、ISO、NIST) | 哪些身份保证和认证级别适用于金融场景 | 独立测试结果、事件类型划分、跨司法管辖区比较 |
该主题产生的产出
- Onboarding risk guides for remote and digital channels
- 关于合成身份、骡子账户和 深度伪造
- 关于分层尽职调查与包容性的说明
- 国家数字身份保证等级与金融开户的对照
专家提出的问题。
- 数字化开户和验证方法如何在安全、便利与包容之间取得平衡?
- 合成身份和账户接管是如何产生的,又如何被检测?
- Who is excluded by identity requirements, and why?
- 数字身份 系统 如何保护隐私?
- How does risk-based customer due diligence work in practice?
Verification, fraud and exclusion.
Know-your-customer requirements oblige financial institutions to identify and verify customers and assess the risk they pose. Digital onboarding uses document scanning, biometric matching, database checks and device signals. Strength of verification varies, and attackers exploit the weakest methods. Synthetic identity fraud combines real and fabricated information to create identities that pass checks; account takeover uses stolen credentials or social engineering; and generative AI enables convincing fake documents and voice or video impersonation.
Exclusion is the other side of the problem. Many people lack formal identity documents or proof of address, and rigid requirements can deny them access to accounts, credit and payments. Risk-based approaches allow simplified due diligence for lower-risk products, and national digital identity systems can widen access, but they raise privacy and surveillance concerns and can exclude those not enrolled.
Identity is also a systemic dependency: when a widely used identity provider fails or is compromised, onboarding and authentication can stop across many institutions.
Recovery of identity after compromise is an under-studied problem. Victims of identity theft may spend long periods restoring access and credit standing, and systems designed for onboarding often lack equally strong processes for re-establishing a person’s identity.
方法及其局限。
Fraud pattern analysis
Studying attack methods and detection; data are fragmented across firms.
访问研究成果
Measuring who is excluded and why; under-documented people are hard to reach in surveys.
Privacy engineering
评估选择性披露等保护隐私的设计。
局限:欺诈数据较为敏感且定义不一致,身份规则对普惠性的影响也难以与其他障碍区分开来。
问题与解答。
What is risk-based customer due diligence?
What is a synthetic identity?
数字身份能否同时减少欺诈和排斥?
Fintech Nexus 是否核验身份?
GRA 是否核实身份?
Can identity rules be both strict and inclusive?
生物特征验证安全吗?
从本页贡献证据。
Reports 是联合体的季刊。您对该主题的了解可以成为下一期中的证据、案例或更正,署名归您,并在属于您本人作品时以 CC BY 4.0 发布。
- 电子邮件门户 GW-GRA从本页发送证据、更正、案例、方法、审查、翻译或概况投稿。它会带上本页标签送达,由负责的编辑接收。
- ZenodoOpen depositDeposit it and submit to the communityPublish your own record on Zenodo with your ORCID iD, then submit it to the Reports community there. Curators review it; the DOI stays yours.
- 报告全球风险联盟(Global Risks Alliance)查看所有贡献方式贡献类型、各自的评审方式、您获得的署名、许可协议,以及绝不接受的内容。
切勿发送个人数据、敏感基础设施细节或您无权分享的任何内容。贡献在使用前会经过审查;发送贡献不赋予署名权或编辑权,赞助也绝不能换取研究结论。
法律文件。
本网站上的地图、数据和图表仅作为背景和一般信息展示。它们不是建议、预测或预警;信息的局限见免责声明和数据说明。