Because many attacks target the factor delivery or the session after the password is accepted. Token theft, prompt bombing, and adversary-in-the-middle techniques can defeat MFA if the second factor is not bound to the right device and session. Effective MFA requires phishing resistance, not just an extra prompt.
Why Weak MFA Still Breaks at the Session Layer
Weak MFA fails because real attackers rarely stop at the password prompt. They target the factor delivery path, the browser session, or the token issued after authentication is complete. That is why push fatigue, intercepted one-time codes, and adversary-in-the-middle relay attacks keep working when MFA is not phishing-resistant. Guidance from CISA cyber threat advisories consistently shows that authentication controls must survive active attack, not just satisfy login policy. NHIMG’s 52 NHI Breaches Analysis also reinforces a broader pattern: when credentials or tokens are exposed, access can be abused long after the initial login event is over.
Security teams often overestimate MFA because they measure enrollment coverage instead of attack resistance. A method can be “multi-factor” and still be weak if the second factor can be replayed, proxied, or socially engineered. In practice, many security teams discover the weakness only after a valid session is already being reused from an attacker-controlled device.
How Attackers Bypass the Second Factor in Practice
Attackers focus on the parts of authentication that are easiest to bend. If a user approves a push notification under pressure, the factor proves only that someone had the phone, not that the login was legitimate. If a six-digit code is read aloud or phished in real time, the code is still valid when the attacker relays it. If the session cookie is stolen after login, MFA may never be challenged again.
That is why strong MFA depends on binding the factor to the correct device, origin, and session. Phishing-resistant methods such as FIDO2/WebAuthn reduce relay risk because the credential is cryptographically tied to the relying party. NIST SP 800-53 Rev. 5 emphasizes stronger authentication controls, and current guidance suggests pairing MFA with device posture, token binding, and short session lifetimes rather than treating MFA as a standalone gate. See also Top 10 NHI Issues for the broader identity abuse patterns that show up once access is granted.
- Push approvals are vulnerable to fatigue and coercion.
- OTP codes can be intercepted, proxied, or replayed in real time.
- Browser sessions can be hijacked after successful MFA.
- Help-desk recovery flows often bypass stronger authentication altogether.
For defenders, the practical test is simple: if an attacker can capture, relay, or reuse the factor without possessing the bound device and session context, the MFA implementation is not phishing-resistant. These controls tend to break down in remote access environments that still rely on legacy OTP, broad session durations, and weak recovery workflows because the attacker only needs one successful relay to win.
What Strong MFA Needs to Hold Up Under Attack
Tighter authentication usually increases user friction and rollout complexity, requiring organisations to balance resistance against support burden. The tradeoff is worth it, but the design has to match the threat. Strong MFA should use phishing-resistant authenticators, enforce step-up checks for risky actions, and expire sessions aggressively when risk changes. Where possible, bind the factor to device attestation or hardware-backed keys so a stolen password alone cannot complete the flow. The Anthropic report on AI-orchestrated cyber espionage also shows why interactive abuse matters: attackers increasingly combine automation with social and technical pressure to outpace human response.
Practitioners should also review recovery paths, device registration, and session revocation. A strong login can be undone by a weak reset flow or an overlong refresh token. NHIMG’s Microsoft Midnight Blizzard breach illustrates how identity compromise often extends beyond the initial credential event into tokens, sessions, and downstream access. Best practice is evolving, but the direction is clear: authentication must be resistant to replay, not merely interactive.
In environments with legacy VPNs, shared admin accounts, or SMS-based recovery, these protections often erode because the weakest path becomes the one attackers choose first.
Standards & Framework Alignment
This section maps relevant standards and security frameworks to the operational risks and controls described in this guidance.
OWASP Non-Human Identity Top 10, OWASP Agentic AI Top 10 and CSA MAESTRO address the attack and risk surface, while NIST AI RMF and NIST CSF 2.0 set the governance and control requirements practitioners need to meet.
| Framework | Control / Reference | Relevance |
|---|---|---|
| OWASP Non-Human Identity Top 10 | NHI-01 | Weak MFA often fails where identity proofing and session binding are weak. |
| OWASP Agentic AI Top 10 | Agentic systems inherit weak auth risks when tokens or sessions are reusable. | |
| CSA MAESTRO | MAESTRO addresses identity controls and attack paths in dynamic AI and cloud systems. | |
| NIST AI RMF | AI RMF supports governance of authentication risk in adaptive, high-variance environments. | |
| NIST CSF 2.0 | PR.AA-1 | Authentication and access control are directly implicated by MFA bypass patterns. |
Design auth flows so credentials and sessions cannot be replayed by autonomous workloads or attackers.
Related resources from NHI Mgmt Group
Deepen Your Knowledge
Reviewed and updated by the NHIMG editorial team on August 16, 2026.
NHI Mgmt Group — the #1 independent authority on Non-Human Identity, IAM, and Agentic AI security. nhimg.org
