档案库 · 金融科技 · 战略决策 · 2012–2026
Socure豪赌AI代理战胜AI欺诈；1.56亿美元融资使其估值达52亿美元
身份验证初创公司Socure以52亿美元估值融资1.56亿美元，并收购Fravity以实现欺诈调查自动化，此前一年内其网络上AI制造的欺诈案件激增8000%
Socure
做的是什么生意
Identity verification and fraud-prevention software: AI/ML models that help banks, fintechs and government agencies decide instantly whether a new customer is real, sold on usage- and transaction-based SaaS
启动资金：Over $742M in disclosed funding since its 2012 inception, most recently a $156M growth round led by Summit Partners
起因
Johnny Ayers co-founded Socure in 2012 in Incline Village, Nevada, and still runs it as CEO. The founding bet was that AI and machine learning could beat legacy identity checks: Socure's pitch is helping banks, fintechs and government agencies 'approve real customers instantly while stopping fraud,' sold per use rather than as an annual consulting-style project
经过
By 2026 the bet had scaled: more than 3,000 enterprise customers — 19 of the 20 largest U.S. banks, more than 600 fintechs, major sportsbook and prediction-market operators, and 160 public-sector organizations, including Capital One, Citi, Chime, Robinhood, DraftKings and Revolut — and Socure says it ended the second quarter with $364M in annual recurring revenue, up 63% year over year, adding 95 customers such as Circle and MoneyLion in the quarter. Then generative AI became both the problem and the next product: AI-driven fraud across Socure's network jumped 8,000% in a year, while the alerts that automation flags for human review kept piling up — Liminal sizes the financial-crime investigation market at $71.1B, with 53% of banks spending at least an hour per alert. In late August 2026 Socure announced a $156M growth round led by Summit Partners (Goldman Sachs Alternatives, Wells Fargo and Docusign also participated), valuing it at $5.2B, up from $4.5B at its 2021 Series E, and pushing disclosed funding past $742M. The round accompanied the acquisition of Austin-based Fravity, whose AI-native agents automate fraud, risk and compliance investigations; it becomes RiskOS_Agents inside Socure's RiskOS platform, starting with watchlist screening and know-your-business checks. Across existing deployments the companies say Fravity cut cost per case 80%, sped resolution fivefold and cut false positives by up to 70%. Socure also won a five-year, $163M federal contract in May 2026 to provide identity-proofing for Login.gov, and had more than 550 employees as of March 2026
还没有结局，它还在跑。
背景
Socure是一家位于内华达州因克莱恩村的公司，向银行、金融科技公司和政府机构销售AI身份验证和防欺诈服务。Johnny Ayers于2012年联合创立该公司并担任CEO；产品承诺是“即时批准真实客户同时阻止欺诈”，按使用量和交易量计费SaaS而非年度项目费。
到2026年，已扩展至3000多家企业客户，包括美国前20大银行中的19家、600多家金融科技公司和160个公共部门组织；Socure称第二季度年经常性收入达3.64亿美元，同比增长63%，当季新增95家客户。它还报告AI驱动的欺诈在其网络上一年内激增8000%，因为生成式AI让伪造可信身份变得廉价且易于自动化攻击。
Socure在2026年8月下旬的应对措施是获得由Summit Partners领投的1.56亿美元增长轮——公司估值达52亿美元，高于2021年E轮的45亿美元，披露融资超过7.42亿美元——同时收购了位于奥斯汀的Fravity，其AI代理自动化欺诈、风险和合规调查。Fravity将作为RiskOS_Agents在Socure的RiskOS平台内运行，从名单筛查和KYB检查开始；两家公司称其现有部署将单案成本降低80%，解决速度加快五倍。
这些举措是在更广泛扩展之后：2026年5月，Socure赢得了一项为期五年、价值1.63亿美元的联邦合同，为Login.gov提供身份验证，截至2026年3月员工超过550人。Liminal将金融犯罪调查市场估为711亿美元，53%的银行每次警报至少花费一小时——这正是Fravity收购过来要自动化的积压工作。
这件事要成立，得有什么
- 身份检查对银行来说是经常性、受监管且高量的成本，因此当客户处理更多人员和交易时，提供实时验证的供应商会随之增长
- 生成式AI使最初的赌注更为有力：Socure网络上AI驱动的欺诈一年内上升8000%，身份验证成了决定谁能在线开展业务的关卡
- Fravity瞄准了欺诈中最劳动密集的部分——Liminal估计调查市场达711亿美元，53%的银行每次警报花费超过一小时——并声称能将单案成本降低80%
- 收购已在共享企业客户中使用的代理平台，让Socure能立即推出RiskOS_Agents，而非从头构建调查工作流软件
- 一份为Login.gov提供的1.63亿美元联邦合同表明同一身份产品也销售给政府，拓宽了银行和金融科技之外的客户基础
可借鉴之处
助长欺诈者的AI同时也创造了对验证的需求；制胜之策是收购能够将随之而来的调查积压转化为产品的代理工具
后续进展
截至2026年9月5日，Socure仍在运营并扩张：估值52亿美元，来自Summit Partners领投的1.56亿美元轮，年经常性收入3.64亿美元同比增长63%，拥有3000多家企业客户，Fravity正整合进RiskOS作为RiskOS_Agents。尚无最终结局——新赌注，即AI代理能大规模自动化欺诈调查，才刚刚开始，而关于单案成本降低80%和速度提升五倍的说法来自Fravity早前部署的数据。
资料来源
- Socure Secures $156M at $5.2B Valuation, Acquires AI Fraud Investigation Startup Fravity
- The Week's 10 Biggest Funding Rounds: AI Tools And Assistants Lead Sparser Lineup Of Megadeals
发现哪里写错了？告诉我们。
轮到你了
你刚读完一家。说说你在做什么，看看谁在赌同一件事。
免费账号 · 3 次免费提问 · 不用绑卡