A newly granted patent to Theta Lake highlights a growing compliance challenge in modern digital workplaces: identifying which applications appear on screen during collaboration sessions.
According to Theta Lake, as organisations increasingly rely on tools such as video conferencing and messaging platforms, the visibility of sensitive applications during screen sharing has become an overlooked but significant risk.
The patent, issued in the United States as US 12,464,032, outlines a system designed to visually identify applications displayed during electronic communications. By analysing video and screen-share content, the technology can determine which software interfaces appear during meetings or in shared visual content. The development strengthens Theta Lake’s wider portfolio of intellectual property focused on video compliance and oversight of digital communications.
The patent builds upon the company’s broader approach to compliance in modern collaboration environments. Previous innovations from the firm have focused on context-based policy detection across spoken dialogue, shared visual material, and digital communications within video meetings.
Other patents have addressed areas such as participant identification and AI-driven review workflows. With this latest development, the technology aims to add another layer of visibility by pinpointing precisely which applications were displayed during shared sessions.
Screen sharing has become a central feature across collaboration platforms such as Zoom, Microsoft Teams and Webex. Despite its widespread use, it often receives limited oversight from compliance teams. When employees share their screens during meetings, every open window or application visible on their display can potentially be seen by meeting participants. If sessions are recorded, that visibility extends even further to anyone reviewing the recording later.
This can create unexpected risks. A shared screen might reveal a CRM interface containing client data, a spreadsheet displaying financial or salary details, or a development environment showing internal code or infrastructure information. Even background applications or partially visible windows can expose sensitive information without the presenter realising it.
The same risks can also arise outside of live meetings. Screenshots shared through messaging platforms or attached to emails may expose sensitive data in much the same way as a screen share. For organisations responsible for monitoring communications and safeguarding confidential information, identifying these moments is critical.
Historically, detecting such incidents has relied on manual review processes. Compliance teams would need to watch recorded meetings and visually inspect frames in order to determine whether sensitive applications appeared during a session. Given the sheer volume of digital communication in modern enterprises, that approach is both time-consuming and difficult to scale.
The system described in patent 12,464,032 introduces an AI-driven method to analyse visual content across collaboration channels. The technology reviews screen shares, webcam feeds and shared digital whiteboards to automatically detect when specific applications appear on screen.
At the core of the approach is what the system describes as an application fingerprint. This fingerprint acts as a signature for a particular software interface, combining both textual and visual indicators. Text-based signals may include menu labels, toolbar names, URLs or other interface wording. Visual cues could involve logos, button layouts, form fields or the grid structure typically associated with spreadsheet software.
Machine learning models analyse these characteristics to recognise and distinguish different applications. The models are trained on large collections of application screenshots to learn the distinctive patterns associated with each interface.
The system is capable of identifying a wide range of application types. These include office productivity tools, CRM and HR platforms, financial systems, development environments and email clients. These are precisely the types of applications most likely to display sensitive information such as personally identifiable information (PII), confidential corporate data or financial records.
Detection can also draw on contextual signals beyond visual analysis alone. For example, if a meeting participant states during a conversation or in chat that they are about to share a specific application, that information can be used to improve the system’s identification accuracy.
Automating the detection process provides a key advantage for organisations operating under strict compliance and recordkeeping requirements. Human reviewers analysing hours of recorded meetings may overlook critical frames due to the scale and speed of modern communication. An automated system, however, can apply the same detection process consistently across every frame of every session.
For regulated organisations, this consistency provides a more reliable way to determine whether sensitive applications were visible during a meeting or communication event. Instead of relying on manual inspection, organisations can generate auditable records showing when and where such applications appeared.
As collaboration tools continue to reshape the way businesses communicate, the ability to monitor visual content within digital interactions is becoming increasingly important. Technologies designed to detect and analyse shared applications may therefore play a growing role in helping organisations manage compliance risks across modern workplace communication channels.
Copyright © 2026 RegTech Analyst
Copyright © 2018 RegTech Analyst
