Loading
Loading
We use strictly necessary cookies to run this site, and analytics cookies to understand how it's used. See our Privacy Policy for details.
Tom Uren and Amberleigh Jack talk about new research that shows that Chinese open-weight AI models can be reeducated to remove pro-CCP bias.
Summary
Good news for American businesses that want cheap AI, but we suspect that many knives will still be out for Chinese open-weight models. They also discuss the near-misses from the PHP and Brotli bugs. We collectively dodged the bullet so that we only got grazed, but the good luck won’t last forever.
Finally, Tom looked at CISA’s 2030 strategy. Amazingly, it doesn’t really mention AI at all. Bizarre!
This episode is also available on YouTube.
KazaSec's take
AI-related security incidents are a genuinely new category, prompt injection, model manipulation, and data leakage through an LLM integration don't map cleanly onto traditional application security testing, and are worth assessing deliberately rather than assuming existing controls already cover them.
Coverage details
We've archived 1191 other articles touching the same topic (amberleigh, risky business, about) , see the full security news archive.
Related security advisories
Relevant from KazaSec
More coverage on this topic
We help organizations find and fix the gaps before they make headlines.