Webinar: 'Invisible Generative AI Risks: How IT Departments Should Control Them'

The Reality of Field-Led Generative AI Usage Outpacing Existing SASE As the business application of generative AI expands rapidly, field-led adoption is leading the way in many companies. While tools like ChatGPT are easy to integrate into workflows, they are being used in ways unanticipated by traditional security designs, often exceeding the capabilities of existing SASE environments. With AI usage expected to grow further, IT departments are pressured to decide 'what to allow and how to control it.'

The Challenge of Unknown Input Content and Undetectable Information Leakage A major issue in generative AI usage is the inability to track what information employees are inputting. Despite efficiency gains, the risk of unintentional input of confidential or personal information persists as an 'invisible' threat. Existing SASE provides limited visibility and control, often restricted to a binary 'allow or block' choice, which diverges from practical operations. Consequently, many IT departments continue operations without clear criteria for risk tolerance and control.

Practical Approaches to Visualize and Control AI Usage While Retaining Existing SASE This seminar addresses these challenges by explaining how to visualize and control generative AI usage without significant changes to existing SASE environments. We will introduce methods using Check Point Browser Security to understand actual AI usage starting from the browser and apply flexible policy controls based on risk. Moving beyond the 'allow/deny' binary, we organize concepts for ensuring security without stopping AI usage, providing IT departments with the criteria needed for decision-making.

Recommended For: - IT departments concerned about how to control expanding generative AI usage. - Those facing challenges with undetectable information leakage risks due to unknown input content in ChatGPT. - Those who feel current SASE is insufficient for visualizing and controlling AI usage. - Those considering practical control methods without halting AI utilization. - Those looking to achieve AI compatibility through additional measures without replacing SASE.

Organizers/Co-organizers Check Point Software Technologies Ltd. ■ Cooperation Open Source Utilization Institute Co., Ltd. Majisemi Co., Ltd.

Majisemi will continue to hold webinars that are 'helpful to participants.' Public materials from past seminars and other current recruitment seminars can be found here.

Majisemi Co., Ltd. 3F Shiodome Building, 1-2-20 Kaigan, Minato-ku, Tokyo 105-0022 Contact: https://majisemi.com/service/contact/