Uploaded June 2026 | Updated September 2026, 11 hours ago
In this short interview from Bosch ConnectedWorld 2026 in Berlin, I speak with Prof. Haya Schulmann about one of the most important questions in AI today: what happens to cybersecurity when AI becomes more autonomous?
AI is already changing cybersecurity in two very different ways. It can strengthen defense by helping automate alerts, support analysts and increase security coverage. At the same time, it can also strengthen attackers by lowering the skills needed to create convincing phishing messages, automate attacks or develop malware.
The conversation becomes even more urgent when we bring agentic AI into the mix.
Traditional AI tools may give you the wrong answer. Agentic AI systems can take the wrong action. They can send emails, call systems, change privileges or execute tasks that have consequences in the real world.
That connects directly to one of the big themes at Bosch ConnectedWorld 2026: physical AI. As AI moves into factories, vehicles, robots and connected industrial systems, trust and safety become essential. Security cannot be an afterthought.
In this video, we discuss:
🤖 How AI changes both cyber defense and cyber offense
🎯 Why AI can lower the barrier for cyberattacks
⚙️ What makes agentic AI more risky than traditional chatbots
🔐 Why organizations need strong guardrails for AI agents
🛑 The importance of monitoring, privilege limits and a stop switch
🏭 Why cybersecurity is critical as AI moves into the physical world
#Sponsored #AI #Cybersecurity #AgenticAI #PhysicalAI #BoschConnectedWorld #BCW26 #AISafety #TrustworthyAI #IndustrialAI #DigitalTransformation
In this short interview from Bosch ConnectedWorld 2026 in Berlin, I speak with Prof. Haya Schulmann about one of the most important questions in AI today: what happens to cybersecurity when AI becomes more autonomous?
AI is already changing cybersecurity in two very different ways. It can strengthen defense by helping automate alerts, support analysts and increase security coverage. At the same time, it can also strengthen attackers by lowering the skills needed to create convincing phishing messages, automate attacks or develop malware.
The conversation becomes even more urgent when we bring agentic AI into the mix.
Traditional AI tools may give you the wrong answer. Agentic AI systems can take the wrong action. They can send emails, call systems, change privileges or execute tasks that have consequences in the real world.
That connects directly to one of the big themes at Bosch ConnectedWorld 2026: physical AI. As AI moves into factories, vehicles, robots and connected industrial systems, trust and safety become essential. Security cannot be an afterthought.
In this video, we discuss:
🤖 How AI changes both cyber defense and cyber offense
🎯 Why AI can lower the barrier for cyberattacks
⚙️ What makes agentic AI more risky than traditional chatbots
🔐 Why organizations need strong guardrails for AI agents
🛑 The importance of monitoring, privilege limits and a stop switch
🏭 Why cybersecurity is critical as AI moves into the physical world
#Sponsored #AI #Cybersecurity #AgenticAI #PhysicalAI #BoschConnectedWorld #BCW26 #AISafety #TrustworthyAI #IndustrialAI #DigitalTransformation
