Recent discussions on Hacker News highlight significant security and usability concerns regarding Atlassian’s AI tool, **Rovo**. **Security Vulnerabilities:** Critics point to critical flaws in Rovo’s design, specifically its susceptibility to "prompt injection" attacks. Because the system can dynamically construct and open URLs without sufficient guardrails, researchers warn that it can be manipulated to exfiltrate sensitive data. Despite these known risks, some users feel Atlassian has been slow to implement effective mitigations. **Usability and Privacy:** Beyond security, many users are frustrated by the intrusive rollout of Rovo, which has been integrated into Jira and Confluence interfaces without clear opt-out options for individual users. Community sentiment is largely negative, with many citing the AI’s tendency to produce "slop," hallucinate, and perform redundant tasks like bloating text instead of summarizing it. **Data Practices:** The discussion also reignited concerns over Atlassian’s data policies, specifically its default opt-in for customer data to be used in model training. This has led some organizations to re-evaluate their reliance on Atlassian, with many arguing that the added AI features are a downgrade that compromises both privacy and workflow efficiency.