A malicious user could craft input that is stored in conversation memory and later interpreted by the model in an unintended way. Applications using the affected advisor with user-controlled input may be susceptible to manipulation of model behavior across conversation turns.
A malicious user could craft input that is stored in conversation memory and later interpreted by the model in an unintended way. Applications using the affected advisor with user-controlled input may be susceptible to manipulation of model behavior across conversation turns.
为什么是这个 VPI(可解释·实验性)
VPI 计算依据
| 影响度 | 82.00 |
| 利用信号(无额外利用信号) | ×1.00 |
| VPI | 82.00 |
VPI 公式 vpi-v1
| 来源 | CVSS 版本 | 基础评分 | 严重程度 | 向量字符串 | 评估日期 |
|---|---|---|---|---|---|
| OSV3rd | 3.1 | 8.2 | HIGH | CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:H/A:N | 2026. 05. 19. |
| NVDNIST | 3.1 | 8.2 | HIGH |
| CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:H/A:N |
| 2026. 05. 13. |