A dark side to LLMs. [Research Saturday]

CyberWire Daily

N2K Networks, Inc.

Technology, Tech News, Daily News, News

4.8 • 1.1K Ratings

🗓️ 8 April 2023

⏱️ 17 minutes

🔗️ Recording | iTunes | RSS

🧾️ Download transcript

Summary

Sahar Abdelnabi from CISPA Helmholtz Center for Information Security sits down with Dave to discuss their work on "A Comprehensive Analysis of Novel Prompt Injection Threats to Application-Integrated Large Language Models." There is currently a large advance in the capabilities of Large Language Models or LLMs, as well as being integrated into many systems, including integrated development environments (IDEs) and search engines. The research states, "The functionalities of current LLMs can be modulated via natural language prompts, while their exact internal functionality remains implicit and unassessable." This could lead them to be susceptible to targeted adversarial prompting, as well as making them adaptable to even unseen tasks. Researchers demonstrated these said attacks to see if the LLMs needed new techniques for more defense. The research can be found here: More than you've asked for: A Comprehensive Analysis of Novel Prompt Injection Threats to Application-Integrated Large Language Models Learn more about your ad choices. Visit megaphone.fm/adchoices

Transcript

Click on a timestamp to play from that location

0:00.0	You're listening to the CyberWire Network, powered by N2K.
0:07.0	Today's episode is sponsored by SRM, your first call for cybersecurity and
0:18.1	investigations. Threats today are evolving faster than ever before and since 2005 SRM has pioneered
0:25.3	tailored security solutions for global corporations and their executives.
0:29.5	Whether it's defending against cyber attacks with their award-winning team of ethical hackers and incident response specialists,
0:36.4	or navigating the murky waters of compliance and ESG challenges,
0:40.9	SRMs, Insight and Straight straightforward advice will help you navigate complex risks
0:46.4	and emerge more resilient.
0:48.4	Their secret, a culture that nurtures the sharpest minds, giving them access to the newest technologies and the freedom
0:55.3	to solve problems in new ways, enabling them to craft simple effective solutions for your
1:01.4	unique cyber challenges.
1:03.7	Search your first call to discover how SRM can help your business. Hello everyone and welcome to the CyberWire's research Saturday.
1:27.0	I'm Dave Bitner and this is our weekly conversation with researchers and analysts tracking down the threats and vulnerabilities,
1:34.1	solving some of the hard problems of protecting ourselves in a rapidly evolving cyberspace.
1:39.7	Thanks for joining us. We all have seen the news after Charyngue T.
1:48.0	We all have seen the news after Chary-T
1:51.0	and we all have seen how people are starting to get really interested and
1:56.0	hyped by the new technology and that we motivated our colleagues to actually inspire them to think that there might be an issue with this integration.
2:07.0	That's Sahar Abdonabe from Sisba Hemholz Center for Information Security.
2:13.4	The research we're discussing today is titled,
2:15.8	A Comprehensive Analysis of Novel Prompt Injection Threats
2:19.5	to Application Integrated Large Language Models. integrated large language models. There might be some new security vulnerabilities that we are really not noticing when we put these
	...

Please login to see the full transcript.

Previous episode | Next episode

Disclaimer: The podcast and artwork embedded on this page are from N2K Networks, Inc., and are the property of its owner and not affiliated with or endorsed by Tapesearch.

Generated transcripts are the property of N2K Networks, Inc. and are distributed freely under the Fair Use doctrine. Transcripts generated by Tapesearch are not guaranteed to be accurate.