BYTES | Agentic AI Versus Generative AI

Episode Summary

In this “Byte,” Akvile Jaseviciute from Skadden‘s IP and Technology team breaks down two key AI terms: generative AI and agentic AI. Generative AI models like ChatGPT, Gemini and DALL-E create new content such as emails, images and summaries. Agentic AI, on the other hand, refers to emerging systems that are designed to plan, make decisions and carry out actions with limited human input. As AI systems start to make decisions and operate independently, businesses face tough questions about ownership, liability, bias, data protection and IP.

Voiceover (00:01):
Welcome to Bytes from SkadBytes, jargon-free bite-size insights from Skadden's IP and Tech team on the key issues shaping the tech landscape.

Akvile Jaseviciute (00:13):
Hi. I'm Akvile Jaseviciute from the IP and Technology team here in Skadden London. In this Bytes, we're breaking down two big AI terms you've probably already heard of, generative AI and agentic AI, and what they really mean in practice. Generative AI is what most people think when they hear AI. These are models like ChatGPT, Gemini or DALL-E, designed to create new content such as emails, images, summaries and even music. Their core function is generation. They learn from vast amounts of data to create new content that resembles the data they were trained on. This is great for automation of written tasks such as translation or various communications. Agentic AI, on the other hand, refers to emerging systems designed not to just generate content, but to plan, make decisions and carry out actions with limited human input. Tools like Microsoft AutoGen or CrewAI are designed to complete complex tasks by breaking them down into steps, making decisions, and then acting on them.

(01:14):
Agentic systems often build on top of generative models, but then add orchestration layers that allow the AI to reason, plan and act across multiple steps or tools. With added capability comes greater complexity and greater legal risk. As AI systems start to make decisions and operate independently, businesses face tough questions on ownership of generated content, liability for mistakes or faulty outcomes, integrated bias, data protection and IP. This is particularly relevant where AI is used in HR, finance, legal advice or other high-impact functions regulated under the EU AI Act. Broad application of AI across multiple work streams, especially if they involve autonomous decisions affecting people or critical systems, may trigger stricter obligations under the EU AI Act. As a key takeaway, businesses should always consider the impact on societies and individuals before embedding AI across their operations.

Voiceover (02:12):
Thanks for listening to Bytes. Be sure to subscribe for more tech insights. Additional information about Skadden can be found at Skadden.com.