Generative AI is on the rise, but with it comes concerns about data privacy and security. According to a new report from Gartner, over a third of organizations adopting generative AI are investing in AI application security solutions to protect against data leaks. Additionally, they are planning to increase investments in privacy-enhancing technologies, AI model operationalization, and model monitoring.
The report reveals that many business leaders are enthusiastic about generative AI but also recognize the need for backup options. They want to prevent inaccurate and harmful outputs and the potential for their proprietary information to be leaked through public AI services. One of the biggest concerns is leaked sensitive data in AI-generated code, followed closely by worries about incorrect outputs or biased models.
Avivah Litan, a distinguished VP analyst at Gartner, explains that organizations are wary of trusting public AI options such as Azure and Google with their data. They are concerned about the lack of transparency and liability if a breach occurs. Litan suggests that organizations can mitigate these risks by running on-premise AI models. However, she acknowledges that this may not be feasible for many companies due to resource limitations.
The report also highlights the importance of ModelOps, a governance method similar to DevOps, and privacy-enhancing technologies (PETs). ModelOps enables firms to automate the oversight of AI models and ensure they operate effectively and safely. PETs, on the other hand, protect data while in use through encryption, preventing unnecessary exposure.
Recent incidents at Apple and Samsung have brought attention to the risks associated with generative AI. Apple banned its employees from using certain AI models over fears of sensitive data leakage, while Samsung experienced a source code leak caused by the use of third-party AI. These incidents serve as a reminder of the potential consequences of inadequate data privacy and security measures.
The Gartner survey also reveals an inconsistent view of responsibility for managing the risk of generative AI within organizations. While almost all IT and security respondents agreed that they have a role to play, only a quarter stated that they wholly own this responsibility. IT departments and governance, risk, and compliance departments were identified as the primary entities responsible for overseeing risk management.
In conclusion, generative AI offers exciting possibilities, but it also poses significant challenges in terms of data privacy and security. Organizations are investing in AI application security solutions, PETs, and ModelOps to mitigate these risks. However, there is still a lack of clarity regarding responsibility and accountability. The question remains: who should be responsible for managing the risks associated with generative AI? Share your thoughts and comments below!
IntelliPrompt curated this article: Read the full story at the original source by clicking here a fun game: sprunki horror