| name | redesign-a-high-risk-account-recovery-flow |
| category | web |
| description | Redesign high-risk account recovery through threat modeling, proportionate identity proof, anti-coercion controls, accessible alternatives, session and asset protection, and monitored rollout. Use when recovering access could expose money, health, identity, administration, or other high-impact capabilities. |
redesign-a-high-risk-account-recovery-flow
Recovery must not become a weaker alternate login or a permanent exclusion path.
When to use
- Use for financial, health, government, enterprise-admin, marketplace, custody, creator, or other high-impact accounts.
- Obtain qualified security, identity, fraud, privacy, accessibility, legal, support, and safeguarding review.
Preconditions
- Establish product, identity, fraud, security, privacy, accessibility, legal, support, safeguarding, and high-risk-action authority.
- Define account types, assets, roles, threat actors, recovery reasons, jurisdictions, user populations, and loss thresholds.
- Separate recovery approval from any employee or vendor who can benefit from the outcome.
Procedure
Complete identity and threat analysis, recovery proof and inclusive access, high-risk action protection, and rollout and monitoring before broad release.
- Build a recovery threat and control register for lost devices, compromised email, SIM swap, stolen documents, malware, coercion, abuse, insider action, social engineering, synthetic identity, deceased users, and organizational disputes.
- Model existing authenticators, passkeys, devices, recovery codes, trusted contacts, enterprise admins, identity records, transaction history, and prior recovery evidence without treating any one weak channel as decisive.
- Create risk tiers from account capability, asset value, recent changes, device and network evidence, recovery channel integrity, attack signals, and user vulnerability.
- Offer multiple proportionate proof paths with accessible, language-aware, low-connectivity, disability, changed-name, displacement, and domestic-abuse alternatives.
- Minimize document and biometric collection; define purpose, vendor, retention, deletion, human review, appeal, and failure handling.
- Detect and fence compromised email, phone, device, session, recovery method, help-desk, and administrator paths.
- Bind the approved recovery to an immutable case and new session epoch. Revoke or review old sessions, tokens, authenticators, API keys, and trusted devices.
- Separate identity recovery from capability recovery. Initial access enters an independently governed quarantine; withdrawals, beneficiary changes, API keys, enterprise ownership, and other high-impact powers require their own cooling period and reauthorization gate even when authentication recovery succeeds.
- Protect users facing coercion or abuse through safe contact, discreet notices, trained escalation, and controls that do not alert the adversary.
- Require independent review for high-value, conflicting, repeated, employee-assisted, or uncertain recoveries and record the reasoning.
- Test genuine and adversarial journeys across devices, assistive technology, languages, connectivity, channel loss, support, organization roles, and attack campaigns.
- Roll out by account and risk cohort with abuse, exclusion, abandonment, time, appeal, takeover, loss, support, and accessibility monitoring.
- Provide rapid containment and reversal of recovery-created access while preserving legitimate user assets and evidence.
- Review decisions for calibration, disparate exclusion, reviewer drift, vendor error, employee misuse, and changing attacker behavior.
Failure plan
- If identity evidence conflicts, keep high-risk actions restricted and route to independent review rather than guessing.
- If all normal channels are compromised, use a protected exceptional process with stronger review and explicit delay.
- If an abusive person may monitor a channel, do not send notices or instructions through it.
- If the new flow increases takeover or wrongful exclusion beyond thresholds, stop the cohort and restore the safer governed path.
Worked example
A global financial platform redesigns recovery for users who lost devices while attackers use SIM swaps, compromised email, stolen identity documents, help-desk social engineering, and coerced access, and legitimate users include displaced people, changed names, disabled users, survivors, enterprise administrators, and accounts holding high-value assets. The platform tiers risk, offers inclusive proof paths, fences compromised channels, binds recovery to a new session epoch, delays high-risk actions, and independently monitors both takeover and exclusion.
Done
- A recovery threat and control register verifies account capabilities, threats, channels, proof paths, accessibility, safeguarding, ownership, and decision rules
- An identity and account-state validation report proves case identity, evidence, channel integrity, session epoch, authenticators, devices, roles, assets, restrictions, and reviewer decision
- An abuse, accessibility, and rollout report demonstrates adversarial tests, inclusive journeys, high-risk action controls, cohort thresholds, appeals, incident reversal, and calibrated monitoring