Dealing with increasingly complicated agents

Dealing with increasingly complicated agents

Author: Practical AI LLC October 16, 2025 Duration: 54:56

As AI systems move from simple chatbots to complex agentic workflows, new security risks emerge. In this episode, Donato Capitella unpacks how increasingly complicated architectures are making agents fragile and vulnerable. These agents can be exploited through prompt injection, data exfiltration, and tool misuse. Donato shares stories from real-world penetration tests, the design patterns for building LLM agents and explains how his open-source toolkit Spikee (Simple Prompt Injection Kit for Evaluation and Exploitation) is helping red teams probe AI systems.

Featuring:

Links:

Sponsors:

  • Outshift by Cisco - The open source collective building the Internet of Agents. Backed by Outshift by Cisco, AGNTCY gives developers the tools to build and deploy multi-agent software at scale. Identity, communication protocols, and modular workflows—all in one global collaboration layer. Start building at AGNTCY.org.
  • Shopify – The commerce platform trusted by millions. From idea to checkout, Shopify gives you everything you need to launch and scale your business—no matter your level of experience. Build beautiful storefronts, market with built-in AI tools, and tap into the platform powering 10% of all U.S. eCommerce. Start your one-dollar trial at shopify.com/practicalai
  • Fabi.ai - The all-in-one data analysis platform for modern teams. From ad hoc queries to advanced analytics, Fabi lets you explore data wherever it lives—spreadsheets, Postgres, Snowflake, Airtable and more. Built-in Python and AI assistance help you move fast, then publish interactive dashboards or automate insights delivered straight to Slack, email, spreadsheets or wherever you need to share it. Learn more and get started for free at fabi.ai

Upcoming Events: 

  • Join us at the Midwest AI Summit on November 13 in Indianapolis to hear world-class speakers share how they’ve scaled AI solutions. Don’t miss the AI Engineering Lounge, where you can sit down with experts for hands-on guidance. Reserve your spot today!
  • Register for upcoming webinars here!

There's a lot of noise out there about artificial intelligence, but cutting through the hype to find what's genuinely useful can be a challenge. That's the space where Practical AI operates. Hosted by the team at Practical AI LLC, this technology podcast moves beyond abstract theory to explore how AI, machine learning, and large language models are actually being applied right now. Each episode features unscripted conversations with a diverse mix of experts, developers, business leaders, and curious minds. You'll hear tangible discussions about implementing machine learning systems, the realities of MLOps, the evolution of neural networks, and the practical implications of breakthroughs in deep learning and GANs. The dialogue is grounded in real-world scenarios, focusing on how these technologies solve problems, drive productivity, and create value in accessible ways. Whether you're a professional building models, a business person integrating AI tools, or an enthusiast eager to understand the landscape, this podcast offers a clear, conversational entry point. It’s about making sense of a complex field through the lens of practical application, demystifying the concepts that are shaping our world without losing sight of how they work on the ground.
Author: Language: en-us Episodes: 100

Practical AI
Podcast Episodes
Should kids still learn to code? [not-audio_url] [/not-audio_url]

Duration: 39:20
In this fully connected episode, Daniel & Chris discuss NVIDIA GTC keynote comments from CEO Jensen Huang about teaching kids to code. Then they dive into the notion of “community” in the AI world, before discussing chal…
AI vs software devs [not-audio_url] [/not-audio_url]

Duration: 57:00
Daniel and Chris are out this week, so we’re bringing you conversations all about AI’s complicated relationship to software developers from other Changelog pods: JS Party, Go Time & The Changelog.Sponsors:Neo4j – Is your…
Prompting the future [not-audio_url] [/not-audio_url]

Duration: 45:59
Daniel & Chris explore the state of the art in prompt engineering with Jared Zoneraich, the founder of PromptLayer. PromptLayer is the first platform built specifically for prompt engineering. It can visually manage prom…
Generating the future of art & entertainment [not-audio_url] [/not-audio_url]

Duration: 42:06
Runway is an applied AI research company shaping the next era of art, entertainment & human creativity. Chris sat down with Runway co-founder / CTO, Anastasis Germanidis, to discuss their rise and how it’s defining the f…
YOLOv9: Computer vision is alive and well [not-audio_url] [/not-audio_url]

Duration: 42:46
While everyone is super hyped about generative AI, computer vision researchers have been working in the background on significant advancements in deep learning architectures. YOLOv9 was just released with some noteworthy…
Representation Engineering (Activation Hacking) [not-audio_url] [/not-audio_url]

Duration: 43:34
Recently, we briefly mentioned the concept of “Activation Hacking” in the episode with Karan from Nous Research. In this fully connected episode, Chris and Daniel dive into the details of this model control mechanism, al…
Leading the charge on AI in National Security [not-audio_url] [/not-audio_url]

Duration: 52:03
Chris & Daniel explore AI in national security with Lt. General Jack Shanahan (USAF, Ret.). The conversation reflects Jack’s unique background as the only senior U.S. military officer responsible for standing up and lead…
Gemini vs OpenAI [not-audio_url] [/not-audio_url]

Duration: 43:29
Google has been releasing a ton of new GenAI functionality under the name “Gemini”, and they’ve officially rebranded Bard as Gemini. We take some time to talk through Gemini compared with offerings from OpenAI, Anthropic…
Data synthesis for SOTA LLMs [not-audio_url] [/not-audio_url]

Duration: 46:39
Nous Research has been pumping out some of the best open access LLMs using SOTA data synthesis techniques. Their Hermes family of models is incredibly popular! In this episode, Karan from Nous talks about the origins of…

«1...678910