Would a company focused on preventing AI based security breaches be useful?

Joshua Molinare
9 replies

Replies

Krishna Kumar
Using AI to prevent security breaches or preventing security breaches caused by AI? Two different use cases here. The first would be useful. The second, I don't know..
Joshua Molinare
@kkumarkg I defiantly think the first one would have a high chance of success, the second as a business may struggle. If the second project was done with research in mind without worrying about profit I could see it being successful in identifying new forms of AI based security breaches.
Richard Gao
Does AI have a different impact on security breaches? It seems to me that all the security breaches that can be done by AI still fall into general categories of all security breaches, and can be thwarted or prevented in the same ways (password security, separations of duties, etc).
Joshua Molinare
@richard_gao2 I agree, I think a research group focused on possible new attack methods caused by AIs increasing power and ways to prevent them could be useful. For now, I think we are pretty safe!
Amanda Trincher
Of course, yes, now software or AI-based solutions are increasingly appearing on the market and in many industries. All you need is to find an AI consulting company that will test your ideas and create a prototype of such a solution.
Sami Khan
Of course! But there's a lot more to AI security than just breaches - prompt injection, hallucinations, data leakage - and those are just some of the things we are solving at Lakera :)