2869 - Guardium AI Security + watsonx.gov -- 通往可信赖人工智能之路.pdf

《2869 - Guardium AI Security + watsonx.gov -- 通往可信赖人工智能之路.pdf》由会员分享，可在线阅读，更多相关《2869 - Guardium AI Security + watsonx.gov -- 通往可信赖人工智能之路.pdf（30页珍藏版）》请在三个皮匠报告上搜索。

1、Orlando,FLOctober 69IBM TechXchange 20252869Guardium AI Security+watsonx.gov-The road to Trustworthy AIChris Brown,AllTrue.aiMatt Simons,IBMSecuring AI in the Age of Shadow AIAgenda030913172123Meet AlexFinding the AIMitigate vulnerabilitiesContinuous protection DemoWhy it all mattersIBM TechXchange|

2、2025 IBM CorporationMeet Alex the CISOAlex knows there is AI in the organization running in multiple clouds and code repos,embedded in apps and co-pilotsAlexs team has very limited visibility into what exists,where it is,and who is using whatWhile the organization is trying to innovate using generat

3、ive AI,the security team is worried about the risks that come from the data and models in use3IBM TechXchange|2025 IBM CorporationAverage cost of a data breach that included shadow AISource:2025 Cost of a Data Breach ReportIBM TechXchange|2025 IBM Corporation463%$670K97%Lack of AI governance is cost

4、lySource:2025 Cost of a Data Breach ReportShare of organizations that reported an AI-related breach and lacked proper AI access controlsOrganizations with no governance policies to manage AI or prevent shadow AI attacksShadow AI raises the costs of a data breach by USD 200K;higher levels of shadow A

5、I raise it by as much as USD 670KIBM TechXchange|2025 IBM Corporation5One lifecycleSecure+MonitorGovernIBM TechXchange|2025 IBM Corporation6Unified AI lifecycleAlexs approachAutomated AI discovery across multiple clouds,code repos,and co-pilotsPosture management and remediation mapped to OWASP Top 1

6、0(and other)frameworksContinual testing and validation with red team policies,prompt protection,observability,and auditIBM TechXchange|2025 IBM Corporation7Risk downResilience upSecurityReduced costs and less time spent ensuring safe AI usageGovernanceFaster audits and fewer blind spots across the e