SNYKREPORTAICode,Security,andTrust:OrganizationsMustChangeTheirApproach56.4%sayinsecureAIsuggestionsarecommon—butfewhavechangedprocessestoimproveAIsecurity.AIcodingassistantshaveachievedwidespreadadoptionamongdevelopersacrossallsectors.However,manydevelopersplacefartoomuchtrustinthesecurityofcodesuggestionsfromgenerativeAI,despiteclearevidencethatthesesystemsconsistentlymakeinsecuresuggestions.Unfortunately,securitybehaviorsarenotkeepingupwithAIcodeadoption.TechnologyorganizationsneedtoprotectthemselvesagainstAIcodecompletionrisksbyautomatingmoresecurityprocessesandinsertingtherightguardrailstoprotectnotonlyagainstbadAIcodebutalsoagainsttheunprovenpercep...
发表评论取消回复