Ads Dawson on developing the OWASP Top 10 for Large Language Models

06/26/24 • 36 min

Ads Dawson, release lead and founding member for the Open Web Application Security Project (OWASP) Top 10 for Large Language Model Applications project, has no shortage of opinions on securing generative artificial intelligence (GenAI) and LLMs. With rapid adoption across the tech industry, GenAI and LLMs are dominating the conversation in the infosec community. But Ads says the security approach is similar to other attack vectors like APIs. First, you need to understand the context of AI-related vulnerabilities and how an attacker might approach hacking a particular AI model.

In the latest episode of WE’RE IN!, Ads talks about including threat modeling from the design phase when integrating GenAI into applications, and how he uses AI in his red teaming and application security work.

Listen to hear more about:

The misuse of AI, such as creating deep fakes for financial gain or manipulating powerful systems like the stock market

The role of governments in securing the AI space and the concept of “safe” AI

How the infosec community can contribute to OWASP frameworks

Listen to hear more about:

The misuse of AI, such as creating deep fakes for financial gain or manipulating powerful systems like the stock market

The role of governments in securing the AI space and the concept of “safe” AI

How the infosec community can contribute to OWASP frameworks

Previous Episode

Kevin Tambascio on balancing security with availability of services in healthcare

Integrating security into the product development lifecycle is a tall order for any industry. It’s particularly challenging for healthcare, with its wide range of critical needs from HVAC systems to medical devices. Kevin Tambascio, director of cybersecurity data and application protection at Cleveland Clinic, juggles the need for constant vigilance and staying updated on fast-moving threats to hospitals.

In the latest episode of WE’RE IN!, Kevin discusses the importance of compliance and risk assessment, noting that while compliance with rules like HIPAA is crucial, it's equally important to pressure test controls against real-world threats. Ransomware targeting hospital data is the primary threat, while phishing and potential abuse of generative AI also pose significant risks.

Listen to hear more about:

The benefits of forming an AI task force to enact safe and responsible procedures while enabling clinicians and researchers to explore AI’s potential
Effectively communicating cyber threats to non-technical staff by relating them to potential impacts on patient safety and business operations
Application security in healthcare; applications often have access to sensitive patient health information and can be potential entry points for cyber threats

Next Episode

Anand Prakash on cloud security startups and next-gen hacking

Anand Prakash on cloud security startups and next-gen hacking

Anand Prakash, founder of startup PingSafe, shares his insights on building a successful cybersecurity business and his experience as a top bug bounty hunter. He emphasizes the importance of fast execution, accountability and learning from mistakes when growing the company acquired by SentinelOne, where he’s now a senior director of product management.

In the latest episode of WE’RE IN!, Anand touches on India's prominence in global tech – particularly in security research and bug bounty programs – and he shares his personal journey into cybersecurity, which began with a curiosity about hacking at a young age in cyber cafes.

Listen to hear more about:

How bug bounty programs have evolved, with companies now more open to ethical hacking due to increased awareness of data breaches
The viability of government efforts to reduce entire classes of vulnerabilities like SQLi
If AI is effective in improving red teaming and bug bounty hunting