Browser-level SaaS and AI security closes the modern visibility gap

At a glance

What this is: Valence Security argues that the browser has become the key control layer for SaaS and AI because it is where authentication, data movement, prompts, and sharing now converge.

Why it matters: This matters because IAM, SaaS governance, and AI oversight all lose critical context once they stop at login, leaving practitioners unable to govern what users actually do in-session.

Context

The security gap is no longer at the perimeter or even in application configuration. It is inside live browser sessions, where users authenticate, paste data into AI tools, move files, and access sanctioned and unsanctioned SaaS services in the same workflow. When control models stop at API posture or identity state, they miss the behaviour that creates exposure.

Browser-level visibility matters because the browser now functions as the practical workspace for SaaS and AI. That makes it an identity, access, and data-handling intersection point, not just a delivery layer. For IAM and NHI programmes, the lesson is that governance must follow the active session, especially where human access and AI usage blend together.

Key questions

Q: How should security teams govern browser-based AI agents in SaaS environments?

A: Security teams should govern browser-based AI agents as runtime actors, not as ordinary users or static integrations. Give each agent a distinct identity, constrain what it can do in-session, and monitor browser, identity, and SaaS logs together. The key control is not just login validation, but continuous authorization of live actions.

Q: Why do API-only controls miss the biggest SaaS and AI risks?

A: API-only controls miss the live behaviour that happens after authentication. They can show configuration and connected applications, but they usually cannot tell whether a user pasted regulated data into an AI prompt, moved files into an unsanctioned tool, or triggered a browser extension that read session content. The missing signal is session context.

Q: What signals show that SaaS governance is not working?

A: Look for delayed offboarding, repeated manual exports, inconsistent access review responses, and inactive accounts that still carry paid licenses. Those signals indicate that entitlement ownership and usage data are not reconciled often enough to support reliable governance.

Q: What should organisations do when browser activity is the main control gap?

A: They should realign controls to the point of use by monitoring session activity, applying data rules during interaction, and extending discovery to unmanaged browser-native tools. If authentication is the only governance checkpoint, users can still move sensitive data through approved access paths in unsafe ways.

Technical breakdown

Why API-only SaaS security misses in-session risk

API-based SaaS security is useful for inventory, permissions, and configuration drift, but it is blind to what happens after authentication. Once a user enters a browser session, the security question changes from who can access an app to how that access is used. Copying data into prompts, uploading files to unsanctioned tools, and interacting with browser extensions all occur at the interaction layer, not the API layer. That is why API-only telemetry often reconstructs incidents after exposure rather than preventing them in motion. The browser becomes the only place where data, intent, and session activity are visible together.

Practical implication: Treat browser telemetry as a complementary control plane, not an optional enhancement, when designing SaaS and AI governance.

How browser sessions change AI governance

GenAI use creates a new exposure surface because prompts are often written directly in the browser, with sensitive content pasted in from other systems. That means data handling, policy enforcement, and human decision-making all occur at the same point of interaction. If governance only checks whether an AI tool is approved, it misses whether the content submitted to that tool is regulated, proprietary, or operationally sensitive. This is also where identity matters: the same authenticated user can behave safely in one session and inappropriately in the next. Browser-level controls create the missing link between access approval and real usage patterns.

Practical implication: Apply policy decisions at the moment prompts are created and data is submitted, not only at tool approval time.

Shadow SaaS and shadow AI start as browser-native activity

Unmanaged SaaS and AI usage often begins with a browser tab, not with a sanctioned integration. Users can authenticate with private email addresses, skip procurement workflows, and install extensions that interact with page content or session data. Traditional discovery workflows struggle here because they are built around registered applications and known connections, while browser-native usage can appear transient and invisible. The technical issue is not just discovery. It is governance over unmanaged interaction paths that create durable access to sensitive data without a formal onboarding event. That makes browser observability a control requirement, not a convenience.

Practical implication: Extend discovery and governance to browser-originated usage paths so shadow tools are visible before they accumulate sensitive access.

Threat narrative

Attacker objective: The objective is to obtain sensitive business data, prompts, or session content through normal-looking browser activity that bypasses API-centric governance.

1. Entry occurs when a user authenticates to SaaS or AI tools in the browser and begins moving data through active sessions rather than through managed application workflows.
2. Credentialed access is then abused through copy, paste, file upload, or browser extension interaction, which can expose regulated or proprietary information to unsanctioned services.
3. Impact follows when sensitive data leaves governed systems without sufficient session-level visibility, leaving security teams to reconstruct exposure after the fact.

Breaches seen in the wild

• SalesBleed Salesforce Agentforce 2026: Three fixed Agentforce flaws let poisoned web leads make AI agents leak CRM data with zero clicks and send phishing under the agent's identity.

Read our 52 NHI Breaches Analysis report for a comprehensive view of breaches impacting Non-Human Identities including AI Agents.

NHI Mgmt Group analysis

Browser-level visibility is now a governance requirement, not a niche detection feature. Security teams can no longer rely on identity state and API posture alone because those controls describe access, not usage. The browser has become the place where access is exercised, data is moved, and policy boundaries are crossed. Practitioners should treat this as a control-plane shift in SaaS and AI governance.

Session behaviour is the new source of truth for exposure decisions. The article reflects a broader pattern we see across identity and AI programmes: static entitlement data is insufficient once work is mediated through live browser sessions. That creates a named problem we would call browser-side governance gap, where approved access and actual risk diverge. The practical conclusion is that decisioning must happen during the session, not after incident review.

AI security and SaaS security are converging at the browser layer. This convergence matters because the same user can move from a sanctioned application into an unsanctioned AI tool without crossing a formal integration boundary. That means governance models need to unify access, data handling, and behavioural context rather than managing them as separate domains. Security leaders should expect browser-level controls to become part of baseline control architecture.

Identity programmes need a stronger link between authentication and in-session policy. IAM still matters, but it no longer closes the loop once a session is established. For organisations with NHI and agentic AI interests, the same lesson applies to machine-initiated workflows that rely on browser-mediated human interaction or delegated use. Teams should re-evaluate where control actually begins and ends.

The market signal is that posture tools alone are no longer enough. The next governance layer is the one that explains what users, extensions, and AI interactions are doing inside active sessions. That will force practitioners to reassess control boundaries across IAM, DLP, CASB-like visibility, and AI governance. The field is moving toward behaviour-aware enforcement at the point of use.

From our research library:

• Only 5.7% of organisations have full visibility into their service accounts, according to the Ultimate Guide to NHIs.
• 1 in 3 organisations encountered suspicious AI agent activity in 2025, and 99.4% experienced a SaaS or AI ecosystem incident.
• Read next: Ultimate Guide to NHIs — Lifecycle Processes for Managing NHIs

What this signals

Browser-side governance gap: security teams are losing control when policy stops at authentication and resumes only after data has already left the active session. That shift forces IAM, DLP, and AI governance to converge around browser-mediated behaviour rather than separate control silos. Practitioners should assume that the highest-risk actions now happen where the browser is already open, not where the app is registered.

The control implication is straightforward: posture management still matters, but it is no longer enough to describe real exposure. Organisations need session-aware controls that can distinguish sanctioned work from shadow SaaS use, shadow AI interaction, and risky browser extension behaviour. Without that layer, zero trust remains a login-time idea rather than an operating model.

For practitioners

• Extend governance to active browser sessions Instrument the browser so you can see copy, paste, upload, share, and extension activity where data actually moves between SaaS and AI tools.
• Classify sensitive prompt content before submission Apply policy checks when users create prompts or paste data into AI interfaces, especially where regulated, proprietary, or internal content may be exposed.
• Inventory shadow SaaS and shadow AI from browser-originated signals Use browser-level discovery to identify unsanctioned applications, private-email signups, and unmanaged extensions before they accumulate durable access.
• Correlate identity, session, and data movement Merge authentication events with in-session activity so security teams can distinguish approved access from risky behaviour in real time.
• Review zero trust boundaries past authentication Test whether your current model stops at login or continues into the session where users actually handle data and interact with AI.

Key takeaways

• Browser-level context is becoming the missing control layer because active SaaS and AI risk now happens inside the session, not just at login.
• API posture alone cannot explain copy, paste, uploads, extension behaviour, or shadow AI use, so incident teams still lack the most relevant evidence.
• Security programmes should move governance to the point of use by combining identity, session telemetry, and data-handling controls.

Key terms

• Browser-layer visibility: Browser-layer visibility is the ability to observe user activity where it actually happens in the web session, including app use, input, consent, and extensions. For AI governance, it becomes the evidence layer that shows what employees used, what data they exposed, and what access they granted.
• Shadow SaaS: Shadow SaaS is the set of unauthorised or unreviewed software-as-a-service tools used outside central security governance. These applications often bypass normal identity controls, making them difficult to inventory, monitor, and harden against credential-based abuse.
• Shadow AI: AI agents, copilots, or connected tools operating without full visibility or governance from security teams. Shadow AI becomes an identity problem when those systems authenticate with unmanaged tokens, service accounts, or OAuth apps that can reach production resources.
• Usage Telemetry: Usage telemetry is activity data that shows whether a user or organisation is actually using a SaaS application or licence. It helps teams distinguish active business value from dormant entitlement, and it is most useful when combined with ownership and lifecycle records.

Deepen your knowledge

NHI Foundation Level course, the industry's only accredited NHI security programme, covers NHI governance, identity lifecycle, and secrets management for practitioners who need to connect identity controls to broader security operations. It helps teams align access governance with the way modern systems and sessions actually behave.