James Kettle Says AI Is Accelerating Security Research, but Humans Remain Essential
· 5:16 · Watch on YouTube ↗
Summary
PortSwigger’s James Kettle says AI can generate novel security research ideas at scale and expose live systems, but humans remain important for analyzing discoveries and identifying new attack classes.
Key points
- Kettle said an AI research system generated 30,000 novel D-Sync ideas, although their average quality was lower than the best ideas.
- He said AI makes it practical to test many unusual ideas at scale, potentially revealing findings a researcher might not have pursued manually.
- Kettle said researchers still need to analyze discoveries, understand root causes, and recognize threats in different conceptual areas.
- He argued that rapidly advancing research could make defense advice outdated quickly, potentially within days.
- Kettle described humans as an amplifier for AI hackers and said researchers remain important to improving AI security products.
Article Details
- Topics
- [00:43] Whether AI can conduct novel security research
- [01:15] AI-driven research pace and large-scale D-Sync idea generation
- [02:55] Human analysis of discoveries and identifying related threats
- [03:45] The continuing role of security researchers alongside AI
- [04:41] Humans as an amplifier for AI hackers