TLDR: A systematic review of 67 studies identified 22 key data risks in Quantum Artificial Intelligence (QAI), categorized into five areas: governance, risk assessment, control implementation, user considerations, and continuous monitoring. The research highlights that QAI inherits complex vulnerabilities from both AI and Quantum Computing, with privacy concerns being dominant. It reveals significant gaps in current literature focusing specifically on QAI’s combined risks and provides a crucial taxonomy to help researchers and practitioners understand and mitigate these evolving threats for more trustworthy QAI systems.
Quantum Artificial Intelligence (QAI) represents a groundbreaking fusion of Artificial Intelligence (AI) and Quantum Computing (QC). This emerging field holds immense promise for advancements, from enhancing privacy through AI-enabled quantum cryptography to securing future digital systems with quantum-resistant encryption protocols. However, this powerful combination also brings with it a complex array of data risks, inheriting vulnerabilities from both AI’s reliance on sensitive datasets and QC’s potential to undermine classical encryption methods.
A recent systematic review, titled “A Taxonomy of Data Risks in AI and Quantum Computing (QAI): A Systematic Review”, delves into these critical challenges. Authored by Grace Billiris, Asif Gill, and Madhushi Bandara from the University of Technology, Sydney, this research highlights that while AI and QC individually pose significant privacy and security concerns, their integration creates unique vulnerabilities that have not been systematically studied until now. These issues directly impact the trustworthiness and reliability of future AI and QAI systems, making their understanding and mitigation paramount.
The study aimed to expand understanding by systematically reviewing 67 privacy and security-related studies. The researchers followed a rigorous methodology, searching across seven major databases to identify, synthesize, and classify data risks and mitigation strategies in QAI literature published between 2020 and 2025.
Unpacking the QAI Data Risk Landscape
The main contribution of this research is a comprehensive taxonomy developed to systematically classify QAI data risks. The review identified 22 key data risks, which are grouped into five overarching categories:
- Governance and Risk Management
- Privacy & Security Risk Assessment
- Privacy & Security Controls Implementation
- User-Centric Privacy Considerations
- Continuous Monitoring and Improvement
The findings reveal that a significant majority of the identified risks, 18 out of 22 (82%), are primarily related to data privacy concerns, such as user consent and control, transparency, and user rights. The remaining 4 (18%) pertain to data security aspects, including weak access controls and general data security risks. This distribution underscores the prevalent focus on privacy in the QAI context.
A notable insight from the review is the current state of research focus. While 34 studies (50.7%) focused on AI-related data risks and 29 studies (43.3%) examined risks in QC contexts, only 4 studies (6%) specifically addressed QAI. This indicates a significant gap in the literature concerning the combined field, despite its compounded risk landscape.
Key Risk Categories Explained
The research paper elaborates on the findings within each category:
Governance and Risk Management: This category highlights critical weaknesses in data governance, regulatory compliance, and organizational alignment with evolving privacy standards. Many organizations struggle to keep pace with the rapid advancements in AI and quantum technologies, often relying on outdated privacy policies that fail to address the nuanced challenges of these new systems.
Privacy & Security Risk Assessment: This area reveals vulnerabilities in how organizations evaluate and manage data risks within QAI systems. A predominant concern is the exposure of personal data during AI model training and inference, often due to inadequate data handling and insufficient anonymization techniques. The complexity and opacity of emerging technologies further complicate effective risk assessment.
Privacy & Security Controls Implementation: The study points to critical deficiencies in the technical safeguards implemented across AI, QC, and QAI systems. A high-priority concern is the prevalence of weak or misconfigured access controls, which can lead to unauthorized data access and undermine overall security postures.
User-Centric Privacy Considerations: This category identifies a persistent erosion of user trust and a lack of meaningful user agency. Users often have limited opportunities to understand or influence how their data is collected and used. The absence of robust informed consent mechanisms is a particularly salient risk, compromising transparency and user control.
Continuous Monitoring and Improvement: The review exposes substantial deficiencies in the ongoing evaluation and adaptation of data protection practices. Many approaches are reactive rather than proactive, with static policies failing to keep pace with dynamic risk landscapes. The lack of real-time data auditing mechanisms hinders the timely detection and mitigation of emerging threats.
Also Read:
- Next-Generation Security: Exploring Advanced Paradigms for Federated Learning
- Agentic AI: A New Era for Adaptive Cybersecurity in Digital Ecosystems
Implications for Trustworthy QAI
The developed taxonomy provides a structured, research-based model to classify data risks in QAI, offering both theoretical and practical implications. Theoretically, it informs the design of privacy and security-enhancing artifacts and provides a reusable classification model for researchers. Practically, organizations can use this taxonomy to develop tailored data risk assessments for QAI systems, evaluate their internal privacy posture, and identify gaps to ensure alignment with evolving data protection regulations.
In conclusion, this systematic review underscores the urgent need for tailored data privacy and security approaches specific to QAI. By providing a structured classification of risks, the research equips practitioners and policymakers with essential tools to understand, assess, and mitigate the complex and evolving data risk landscape of Quantum Artificial Intelligence, ultimately contributing to the advancement of trustworthy AI and QAI systems.
