Serious question, is open source AI actually a security risk or are we overthinking it?

I was sitting at a truck stop in Flagstaff last week, scrolling through a forum about that new open source model that dropped from Meta. On one hand, having the code out there lets a ton of people find bugs and improve it fast. But then I got to thinking about how it also means bad actors can download the same code and weaponize it without anyone knowing. Has anyone else wrestled with whether the benefits of transparency outweigh the risks of giving tools to the wrong people?

2 comments

2 Comments

caleb_stone1mo ago

Security through obscurity never really works out in the end.

luna8911mo agoMost Upvoted

Oh man @caleb_stone my last project was basically a digital "hide the key under the mat" approach, and shocker it didn't end well. I learned the hard way that nobody wants your poorly hidden passwords.