Push fatigue attacks expose the limits of modern IAM controls

At a glance

What this is: This is an analysis of why push fatigue attacks still work even when authentication is modern and credentials are valid.

Why it matters: It matters because IAM teams can overestimate control strength when approvals look legitimate, even though the trust decision has been manipulated rather than bypassed.

Context

Push fatigue attacks are a behavioural identity threat, not a classical authentication break. The attacker already has valid login data and relies on repeated prompts to coax a user into approving access, which means the control failure sits in trust interpretation rather than credential verification.

For IAM and identity governance teams, the problem is that modern authentication flows can look healthy while the approval signal is being coerced. That makes this a useful case study for anyone responsible for human identity controls, especially where push-based MFA is still treated as a strong trust boundary.

The key governance issue is that many programmes assume a legitimate approval equals a legitimate intent. Once that assumption fails, the organisation needs to think about authentication context, user behaviour, and access decision quality together rather than as separate layers.

Key questions

Q: What breaks when push approvals are treated as proof of trust?

A: The trust model breaks because a user can approve access under pressure, confusion, or fatigue without the authentication stack being technically compromised. In that state, IAM still sees a valid response, but the approval no longer proves intent. Organisations should treat approval as one signal among several, not as the final security decision.

Q: Why do push fatigue attacks remain effective in modern IAM environments?

A: They remain effective because modern IAM often verifies credentials and challenge response, but not the quality of the user’s decision. When an attacker already has valid credentials, the gap is behavioural rather than technical. That means strong login infrastructure can still be bypassed through repeated prompting and human exhaustion.

Q: What should security teams monitor to spot push fatigue abuse?

A: Teams should monitor repeated prompt bursts, unusual approval timing, denials followed by eventual acceptance, and authentication activity that is technically normal but behaviourally repetitive. The key is to look beyond successful sign-ins and identify pressure patterns that suggest a user is being worn down rather than intentionally authorising access.

Q: How do number matching, device trust, and adaptive authentication help against push fatigue?

A: These controls raise the cost of coercion by adding friction, context, and stronger proof before access is granted. Number matching reduces accidental approvals, device trust helps tie requests to known endpoints, and adaptive authentication can escalate when behaviour looks repetitive or unusual. Together they make approval less easy to manipulate.

Technical breakdown

Why valid credentials do not stop push fatigue attacks

Push fatigue attacks work because the attacker does not need to defeat authentication infrastructure. They use valid credentials to trigger repeated prompts until the user approves, often through confusion or frustration. From the IAM system’s perspective, the login looks routine: the account is correct, the channel is working, and the approval arrives through an expected path. The weakness is that the system cannot reliably distinguish coerced approval from informed consent when the signal is only a yes/no user response.

Practical implication: treat valid login plus user approval as an incomplete trust signal unless it is paired with stronger context checks.

Why authentication telemetry can look clean during abuse

Traditional IAM telemetry often focuses on indicators such as device, location, velocity, and credential match. Push fatigue attacks can stay within acceptable thresholds, so they do not always trigger obvious anomalies. The issue is not a broken control but a control model that is too shallow for behavioural coercion. When the same user approves many prompts, the data may still resemble ordinary access activity unless the programme specifically models repeated prompt pressure and approval frequency.

Practical implication: add prompt frequency and approval-pattern analysis to existing authentication monitoring.

Why approval-based trust is the weakest link in modern IAM

Many identity architectures still treat successful approval as the final validation event. Once the prompt is accepted, policies execute and the session begins, even if the approval was induced rather than intentional. That design is convenient, but it creates a governance blind spot because the control boundary sits at the user action instead of the underlying assurance state. In practice, the system is assuming that human confirmation is equivalent to security confirmation, which is exactly what fatigue attacks exploit.

Practical implication: move trust decisions earlier in the flow by strengthening the assurance required before an approval can unlock access.

Threat narrative

Attacker objective: The attacker aims to obtain authenticated access by persuading the user to validate a session the attacker already initiated.

1. Entry begins with valid credentials that allow the attacker to initiate repeated authentication prompts without first breaking the login system.
2. Escalation occurs when the user approves the prompt under pressure, converting a legitimate challenge into attacker-controlled access.
3. Impact follows when the approved session starts normally and the attacker inherits authenticated access that appears trustworthy to downstream controls.

Breaches seen in the wild

• SonicWall SSL VPN account compromises 2025: Attackers used valid credentials to log in to more than 100 SonicWall SSL VPN accounts across 16 environments in October 2025.

Read and download The State of NHI & AI Agent Breach Report 2026, covering 200+ breaches impacting Non-Human Identities including AI Agents.

NHI Mgmt Group analysis

Approval is not the same thing as trust: Modern IAM programmes often treat a successful push response as if it were a reliable signal of user intent. That assumption fails when the actor is under prompt pressure, because the approval can be coerced without any compromise of the authentication stack. The implication is that identity governance has to measure trust quality, not just authentication completion.

Contextual intelligence is the missing control plane: Push fatigue attacks expose a gap between authentication telemetry and decision quality. Device match, location match, and valid credentials can all be true while the approval is still unsafe. The decisive issue is whether the programme can interpret repetition, pressure, and human response patterns as risk signals, which makes behavioural context central to IAM governance.

Modern IAM is still too approval-centric: Many organisations have upgraded the front door but left the trust model intact. If approval remains the point where access becomes legitimate, then attackers only need to manipulate the user at that step. Mature identity programmes should view approval as one input, not the validation event itself.

Push fatigue is a governance failure, not just a detection problem: The article shows that traditional MFA assumptions can survive technically while failing operationally. The control gap is not that authentication is absent, but that the system assumes legitimate user action is always a reliable indicator of intent. That is a lifecycle and policy design issue, not merely a monitoring issue.

Named concept, approval signal inflation: Repeated prompts can make a single approval look more credible than it is. That inflation of the trust signal is exactly what fatigue attacks exploit, and it should force practitioners to reconsider which events their IAM programme treats as evidence of assurance. The practitioner conclusion is to govern approval quality as tightly as credential quality.

What this signals

Approval signal inflation: IAM programmes that still equate a prompt response with intent are overvaluing the wrong event. Push fatigue attacks succeed because the trust boundary is placed at the user response, not at the assurance context that should surround it.

Security teams should expect more behavioural abuse where authentication is otherwise strong. The programme response is not to abandon push workflows entirely, but to make them one part of a broader decision chain that includes device trust, rate controls, and contextual escalation.

For practitioners

• Reduce dependence on push-only approvals Shift high-risk sign-in flows away from single-tap acceptance and toward stronger assurance steps that do not rely on user reflex or frustration.
• Tune alerting for repeated prompt patterns Watch for clustered authentication prompts, repeated denials followed by acceptance, and short bursts of approval activity against the same account.
• Add contextual checks before session creation Require device trust posture, certificate-backed identity signals, or adaptive escalation before the approval event can unlock access.
• Limit authentication prompt frequency Apply rate controls and cooldown thresholds so repeated prompts do not become a viable coercion channel.
• Review whether approval is still your trust boundary Reassess policies that treat push confirmation as proof of intent, especially for privileged or sensitive access paths.

Key takeaways

• Push fatigue attacks exploit the gap between a technically valid login and a coerced human approval, which means modern IAM can fail without any infrastructure compromise.
• The article shows that routine-looking authentication telemetry can hide repeated prompt abuse and that approval itself may be the weakest trust signal in the flow.
• Practitioners should reduce reliance on push-only decisions, add contextual assurance, and rate-limit prompts before coercion becomes a practical access path.

Key terms

• Push Fatigue Attack: A push fatigue attack is a social engineering technique that overloads a user with repeated authentication prompts until one is approved. The attack abuses human response patterns rather than breaking the authentication system, which makes the approval event itself the weak point in the access decision.
• Approval Signal: An approval signal is the user response that authorizes an authentication challenge or session request. In modern IAM, it often carries more trust than it deserves, because it can be produced by confusion, annoyance, or pressure rather than by informed intent.
• Contextual authorization: A policy approach that evaluates access using real-time signals such as task, location, device posture, and time. It is more precise than static role assignment because it matches how autonomous agents operate, where intent and risk can change across a single workflow.
• Authentication Telemetry: Authentication telemetry is the record of signups, logins, returning sessions, and other identity events generated by an auth system. It becomes useful when teams translate those events into operational signals for adoption, lifecycle health, and risk review rather than leaving them as raw logs.

Deepen your knowledge

NHI governance, agentic AI identity, and machine identity lifecycle are core topics in our NHI Foundation Level course, the industry's only accredited NHI security programme. If you are building or maturing an IAM programme, it is worth exploring.