"There is a significant danger in using GenAI technologies in organizations"
"There is a significant danger in using GenAI technologies in organizations"
Ishai Ram, Deputy CEO at SELA, spoke at Calcalist's AI conference about how corporate organizations need to implement GenAI tools with caution.
"There is a significant danger in using GenAI technologies in organizations," said Ishai Ram, Executive VP at Sela during Calcalist's AI conference. "We all have tools like Gemini and others that we interact with, but most of these companies warn employees not to expose too much because it can lead to major problems. Shadow AI involves using generic tools freely, and a lot of information is exposed, which creates serious organizational issues including sharing customer data, personal information, code, or strategies, and it might come back to haunt us."
According to Ram, "Everything we feed as prompts are unprotected and the output is also limited. The worst part is that the responsibility for the result falls on us. We might use the AI generated output, and if it's incorrect the organization bears responsibility. There are issues with exposing customer information, unauthorized access to compliance-related areas, various vulnerabilities, and relying on information that is not necessarily reliable.
"Organizations that want to implement GenAI-based systems must do so with corporate tools. We work with tools that are platform-based, communicate at the organizational level, and encrypt data," he said.
"Vertex provides us with a platform for text, audio, and video. It produces GenAI using a model that allows for accurate verification of our organizational data rather than general data. The organization's service system should adapt the GenAI system. This is the major advantage of our tool: that it is trained on our organization's data."
