Data security’s traditional assumptions—knowing where sensitive data resides, which systems control access, and who is accessing it—no longer hold because data is constantly copied across analytics platforms, lakehouses, tests, clouds, and third parties, and is accessed as often by APIs, automated workflows, and AI-driven systems as by humans. This drift correlates with measurable impact: a global average breach cost of $4.99 million, widespread lack of encryption, and frequent AI-related breaches tied to missing AI access controls. Most enterprises already own many point tools (discovery, encryption, masking, tokenization, monitoring, DSPM), but fragmented policy models yield a tool inventory rather than a coherent architecture.
Data Security Platforms (DSPs) aim to unify discovery, classification, protection, access governance, monitoring, and compliance under a single control model, moving enforcement to the data layer so centralized policy follows data access everywhere. Buyer needs are shaped by enforcement overtaking visibility, governance of non-human principals with privileged-user rigor, crypto-agility becoming mandatory, and regulatory divergence forcing residency and sovereignty into design. Market signals show an unsettled category: 21 vendors assessed vs. 13 previously, with varied origins and uneven maturity.
Across platforms, discovery and classification are no longer differentiators, while DSPM is necessary but insufficient—posture scoring is common, but AI store coverage and automated remediation lag. Data-layer enforcement exists but varies; protection still leans toward masking over cryptography, and advanced data-in-use cryptography is rare. Non-human access governance is a major gap, as are crypto-inventory, disruption-free crypto migration, and credible post-quantum implementations. Generative AI is now embedded in products (including RAG and agentic features), but AI-control claims require proof through pilots. Deployment remains mixed (SaaS, hybrid, self-managed) because enforcement must stay near data and sovereignty constraints often prohibit vendor-operated control planes.
See All Locations
See All Locations