
John Israel
CISO -
KPMGI am a business oriented CISO, and IT executive, with over 28 years of international leadership experience across Europe, Asia and the Americas. I believe that combining diverse perspectives with bold, timely decisions can transform the world. I take pride in challenging the status quo and grafting together the goodness of principled thinking, analytical decision-making, and diverse leadership experiences.
Sessions Presenting
Secure and govern custom AI built on Azure AI and Copilot Studio
On Demand
In Chicago + Online
66% of organizations are testing, planning, or already developing GenAI applications, with an average of nearly 14 applications simultaneously. This can introduce new and increasing security threats. In this session, we will demonstrate how Microsoft can help you enable data security and compliance controls for your custom AI apps, manage vulnerabilities in your AI stack, and detect and respond to emerging AI threats such as prompt injections and model theft.
Trustworthy AI: Future trends and best practices
On Demand
In Chicago + Online
Join us for an enlightening session on building and using trustworthy AI applications that prioritize safety, security, and privacy. Discover best practices for AI security and mitigating emerging threats, learn how to develop applications with built-in safety and guardrails, manage compliance, and safeguard sensitive data.
Secure and govern custom AI built on Azure AI and Copilot Studio
On Demand
In Chicago + Online
66% of organizations are testing, planning, or already developing GenAI applications, with an average of nearly 14 applications simultaneously. This can introduce new and increasing security threats. In this session, we will demonstrate how Microsoft can help you enable data security and compliance controls for your custom AI apps, manage vulnerabilities in your AI stack, and detect and respond to emerging AI threats such as prompt injections and model theft.