Deliberately attacking AI systems to find their weaknesses before others do. Testers try to get models to produce harmful content, reveal secrets or take dangerous actions. Labs run it before releasing new models, and it's increasingly automated with AI itself.
133 stories so far in 2026
Stories per month in 2026| Month | Stories |
|---|
| January | 0 |
|---|
| February | 0 |
|---|
| March | 5 |
|---|
| April | 5 |
|---|
| May | 1 |
|---|
| June | 23 |
|---|
| July | 4 |
|---|
| August | 6 |
|---|
| September | 16 |
|---|
| October | 42 |
|---|
| November | 31 |
|---|
| December | 0 |
|---|