Specialist surveillance solution provider Shield closes 2025 with a sharp uptick in industry recognition, underscoring its growing influence in digital communications governance. Gartner has ranked the Tel Aviv–based firm among the top three providers across all six evaluated use cases in its Critical Capabilities for Digital Communications Governance and Archiving (DCGA), while also naming Shield a Visionary in the accompanying Magic Quadrant. The results place Shield firmly among the emerging leaders reshaping surveillance, archiving and oversight as regulatory expectations tighten around multi-channel communications and AI-driven operations.
The momentum is reinforced by Shield’s inclusion in Deloitte’s Technology Fast 500™, reflecting its position among North America’s fastest-growing fintech firms and signalling accelerating enterprise adoption of unified, AI-enabled communications governance.Gartner’s assessment highlights the breadth of that capability: Shield placed second in both the Investigations and User Governance use cases, and achieved its highest scores in the Connectors category, demonstrating the platform’s ability to consolidate and normalise communications data across an increasingly complex ecosystem.
RegTech Insight spoke with Shield executives Alex de Lucena, Director of Surveillance and Governance Strategy, and Tamar Sharir, Chief Product Officer, to understand the dynamics behind the company’s trajectory – and what their customers’ challenges reveal about the direction of compliance technology.
From Fragmentation to Unification
The starting point, Sharir explains, is architectural rather than aesthetic: “Shield, from inception, was built with the vision of providing one unified platform that serves basically all the use cases… from archiving, record management… all the way to surveillance analytics and all additional models.”
This distinction matters in a market long dominated by stitched-together solutions. Legacy providers often grew by acquisition or bolting new capabilities onto ageing archives. Shield takes the opposite approach, designed as a single environment for capture, archiving, supervision, surveillance, voice review and case management.
De Lucena notes that this foundation differentiates the firm at a time when institutions are looking to break the cycle of incremental fixes. “All of our competitors, for the most part, started as one or the other… in our case, we’ve satisfied all of the above from inception.”
The result, he argues, is that customers tend to engage Shield when legacy systems are no longer delivering. “If you’re stuck with a legacy vendor… you’re dealing with too many false positives, the UI hasn’t changed… or you’re on your AI journey… that’s where we’ll be relevant.”
AI as an Enabler, Not a Replacement
AI – particularly its responsible application – is central to Shield’s strategy. But Sharir is quick to emphasise that adoption must mirror the maturity and risk appetite of each institution. “Our approach and our perspective on AI adoption is really to follow the customer pace… we see customers that want only Gen AI… we see others that are more on the deterministic side.”
Shield therefore supports the full spectrum: lexicon-based rules, deterministic machine learning, supervised models and generative AI. Across all of these, Sharir stresses the same principle: “When we innovate… it’s always with a very responsible mindset… everything is clearly marked, all actions are audited, and the most important thing, everything is left for human decision making, we support with our AI.”
De Lucena describes the role of GenAI in practical terms. Rather than replacing surveillance logic, it strengthens existing processes. “We use GenAI to do two things. One is to look for any alerts that we’ve signalled… to see if we missed anything.” This addresses one of the most persistent questions in surveillance: how to identify risk signals that fall short of alert thresholds yet remain material.
The second application focuses on the opposite challenge – noise. “We also use it to weed out irrelevant communications. The work of reducing noise is ongoing… similar to cleaning your apartment… there’s always dust.”
Making Sense of Complex Communications
One of the more difficult areas of communications monitoring is interpreting “trader-speak”: clipped exchanges, jargon-heavy shorthand, and language that is not obviously incriminating when taken sentence-by-sentence. De Lucena notes the limitations of relying on a single analytical method. “A lot of trade risk in the chat rooms isn’t sentence based… it almost looks like code.”
Shield treats trader jargon as another language – highly compressed, context-dependent and often resembling code rather than natural sentences. Instead of relying on any single analytical method, the firm applies a multi-layered approach. Generative AI helps interpret conversations in full context rather than as isolated phrases, reducing false alerts that arise from literal keyword matches.
At the same time, Shield classifies every identifiable entity (such as equities, bonds or derivatives), flags basic trade-related activity to establish contextual relevance, and applies supervised models alongside GenAI to interrogate weaker signals. By combining these techniques and filtering out noise at the end of the process, Shield aims to surface messages that warrant human review even when no single phrase alone would indicate risk.
This approach extends to voice, where Shield uses its own transcription engine across turrets, mobile devices and other audio sources. Sharir and de Lucena both highlight financial-specific transcription and multi-lingual support as emerging necessities for market abuse and conduct detection.
Human Oversight and Workflow Integrity
Despite the industry’s appetite for automation, both executives stress the role of human judgment. Shield emphasises that, regardless of how advanced the analytics become, human judgment remains essential. Analysts must review and decide on every alert; nothing is automatically escalated or actioned without oversight. Effective compliance, the firm argues, depends on maintaining this human decision point. Shield’s role is to curate the cleanest and most relevant set of alerts possible – reducing noise and highlighting meaningful risk signals – so that reviewers can focus on outcomes that genuinely matter.
This approach enables institutions to handle emerging risk categories, including non-financial misconduct and retail-client interactions. “Just the last year, we saw a big push out of the UK around non-financial misconduct… big banks talking about bringing retail under surveillance,” de Lucena notes.
Partnering for Change: The PwC Collaboration
One of Shield’s most consequential developments this year is its global partnership with PwC, which selected one technology provider per vertical – including Shield for communications governance.
For firms decoupling from legacy vendors, the combination of technology, programme management and change-management capability appears to resonate. Sharir adds: “They help them make the case, carry it over, manage the project, and really uplift the control.”
Looking Ahead: 2026 and Beyond
Shield’s 2026 development priorities centre on deepening its AI-driven surveillance capabilities, particularly through its Fortified Surveillance layer, which is designed to detect weak signals and reduce noise. The company plans to introduce additional specialised agents that can handle different contextual tasks while giving customers clearer control over which types of messages can be safely suppressed.
A key forthcoming feature is Surveillance by Example, which allows reviewers to take a sentence that should have triggered an alert, run it through an LLM to identify similar patterns, and convert that insight into a small, deployable model – complete with an approval and testing workflow – rather than relying on new lexicon rules.
Other roadmap items include adding OCR to capture on-screen text, expanding conduct-risk detections, and incorporating smaller, cost-efficient language models to handle general communications while reserving larger models for more complex reasoning. Overall, Shield plans to advance these components incrementally, bringing them together while continuing to refine the user experience and interface.
The combination of analyst recognition, market traction and technology development suggests a DCGA market moving quickly toward consolidation, proactive oversight and AI-augmented operations. Shield’s leadership believes its unified architecture and AI design philosophy place it on the right side of that curve.
Subscribe to our newsletter
