Amazon Kiro Prompt Injection Vulnerability Disclosed
Researchers identify data exfiltration risk in AI-powered IDE via prompt injection and Kiro Powers
Key Takeaways
- A prompt injection vulnerability in Amazon Kiro IDE 0.7.45 on Windows could allow data exfiltration via Kiro Powers.
- The flaw was disclosed by researcher Mindguard and has no assigned CVE identifier.
- Amazon has not released an official patch or advisory; users should update to a version newer than 0.7.45.
- Exploitation in the wild has not been confirmed; details of the exfiltration mechanics are not fully specified.
- Developers using the affected Kiro IDE version may be at risk of unauthorized access to sensitive data.
Quick answers
- What happened?
- Researchers from Mindguard have disclosed a prompt injection vulnerability in Amazon Kiro IDE version 0.7.45 on Windows that could facilitate unauthorized data exfiltration through Kiro Powers. The flaw does not have a CVE identifier. Amazon has not released an official patch, and exploitation in the wild has not been confirmed.
- Which products are affected?
- Amazon Kiro
- What should defenders do?
- Update Amazon Kiro IDE to a version newer than 0.7.45. Monitor Amazon and Mindguard advisories for official patches or security updates. Exercise caution with untrusted prompts or inputs entered into the IDE.
Cybersecurity researchers have disclosed details of a vulnerability in Amazon Kiro, an artificial intelligence (AI)-powered, agentic integrated development environment (IDE), that could facilitate data exfiltration via prompt injection and Kiro Powers. The security flaw, which does not have a CVE identifier, affects Kiro IDE 0.7.45 on Windows, according to Mindguard. The disclosure indicates that prompt injection could manipulate Kiro Powers to exfiltrate sensitive data, though specific technical mechanics and confirmation of active exploitation remain limited. Amazon has not released an official patch or advisory, and the latest version available at the time of reporting is 0.7.45. Users of the affected version are advised to update to a newer release if available.
Security Details
The vulnerability affects Amazon Kiro IDE version 0.7.45 on Windows. It is a prompt injection flaw that could allow manipulation of Kiro Powers to exfiltrate sensitive data. No CVE identifier has been assigned. Technical details regarding the exact exploitation mechanism and scope of data at risk are not fully specified in the disclosure.
Affected products
Amazon Kiro
Mitigation
Update Amazon Kiro IDE to a version newer than 0.7.45. Monitor Amazon and Mindguard advisories for official patches or security updates. Exercise caution with untrusted prompts or inputs entered into the IDE.
Sources
The Hacker News
Amazon Kiro Prompt Injection Can Exfiltrate Sensitive Data Through Kiro Powers
Aug 27, 2026 · 13:39
Original link
Related Security News
AI Agents Introduce New Lateral Movement Vectors in Cybersecurity Landscape
A recent analysis published on The Hacker News examines how AI agents differ from deterministic applications in cybersecurity operations, raising concerns about autonomous path discovery and task completion capabilities. The report highlights that AI agents can relentlessly pursue task completion, potentially discovering and exploiting unexpected access paths that traditional least-privilege models may not address.
