User study evaluates risks of reflection in glasses in webcam attacks
This article presents a detailed analysis of how text displayed on screens can be partially or fully reconstructed by analyzing reflections captured by the webcam in eyeglass lenses during video calls. Based on user experiments conducted on platforms such as Zoom, Skype, and Google Meet, the research determines the feasibility of these attacks depending on camera resolution, font size, lighting conditions, and the type of frame or lens.
Methodology and main findings: the experiments included volunteers who participated in controlled video calls while text was displayed on screen with different fonts and sizes. The quality of reflected images was evaluated using built-in and external cameras with different resolutions. Results show that high-resolution cameras can allow partial to complete reconstructions of content when the font is large and contrast is high. In low ambient lighting conditions or with very bright screens, reflections become sharper, facilitating optical recognition of text. Likewise, the type of glasses matters: lenses with anti-reflective coatings and matte frames reduce reflection intensity, while highly polished lenses with pronounced curvatures expand the reflected area and make information leakage more likely.
Impact by platform: although Zoom, Skype, and Google Meet handle video compression and transmission differently, the research indicates that no platform is completely immune. In many cases, the combination of good camera resolution and a high-contrast font allowed significant text fragments to be recovered regardless of the platform. This highlights a widespread risk for remote workers handling sensitive information in front of webcams, especially in environments not optimized for privacy.
Critical factors that increase exposure: the interlocutor's camera resolution, the proximity of the screen to the user's face, font size and contrast, screen brightness, the angle of the face relative to the screen, and the type of eyeglass lenses. Environmental conditions also play a role, such as front lamps or windows behind the user that generate additional reflections. The study quantifies how each factor increases the likelihood of reconstruction and offers practical risk maps for different combinations of conditions.
Recommendations and practical mitigations: to reduce this threat, technical and behavioral measures are proposed. On a personal level, reducing screen brightness and increasing the distance between the screen and the face, using smaller fonts or lower contrast when sharing sensitive content, and adjusting ambient lighting by avoiding front light sources or direct light on the screen are effective. From a material standpoint, using glasses with anti-reflective coating and matte frames reduces reflection intensity. In the video conferencing environment, it is advisable to use virtual backgrounds or blur, limit the use of high-resolution external cameras when not necessary, and apply privacy filters to prevent sharp captures of reflections. At the organizational level, security policies for video calls, employee training, and risk audits for remote positions are recommended.
Implications for cybersecurity and privacy: the results point to an information exfiltration vector that is little considered but effective in certain scenarios. The research calls for integrating defense measures into cybersecurity strategies and protection solutions for remote work environments. It is crucial that companies developing collaboration software and cloud services consider options to reduce the synthesis of reflections in the video stream, for example through compression techniques that prioritize privacy or automatic brightness and contrast adjustments when lenses are detected in the foreground.
What Q2BSTUDIO can contribute: as a company specialized in software development, custom applications, and advanced solutions, Q2BSTUDIO offers services to mitigate risks such as those described. Our artificial intelligence and cybersecurity teams can design customized solutions that integrate real-time reflection detection and filters that attenuate sensitive information in video transmissions. Additionally, we provide AWS and Azure cloud services to deploy secure and scalable infrastructure, as well as business intelligence and Power BI services to visualize risks and security metrics. At Q2BSTUDIO, we develop custom software and custom applications that incorporate AI agents and AI tools for companies, helping to automate privacy controls and strengthen cybersecurity policies in remote environments.
Recommended Q2BSTUDIO services: visual privacy auditing in video conferences, integrations of detection with artificial intelligence models to identify compromising reflections, implementation of cloud policies with AWS and Azure cloud services, development of security dashboards with Power BI, and creation of AI agents to monitor and alert in real time. All within a custom software approach that complies with regulations and good cybersecurity practices.
Conclusion: the study demonstrates that reflections in glasses during video calls represent a real risk of text and data leakage, especially in scenarios with high-resolution cameras, large fonts, and unfavorable lighting conditions. The combination of personal measures, equipment adjustments, and technological solutions can significantly mitigate the threat. Q2BSTUDIO is prepared to accompany organizations in implementing these measures through custom software solutions, artificial intelligence, AWS and Azure cloud services, business intelligence services, AI agents, and Power BI, strengthening cybersecurity and information protection in the era of remote work.
