Failure patterns

The 14 ways software keeps failing

Most incidents are not new. They repeat a small number of patterns: a retry that double-charges, an agent that obeys instructions hidden in an email, a config change that reaches every server at once. Each pattern below shows how it unfolds, and links to the real incidents behind it and the safeguards that prevent it.

Explore documented incident coverage by pattern

Patterns 1–7

An incident can have more than one pattern.

Patterns 8–14

An incident can have more than one pattern.

Duplicate action after an ambiguous result
A request times out, the caller assumes it failed and retries, and the money moves (or the order ships) twice.
- Design reference: idempotency keys make retried payment API calls safe (Stripe)
CWE-8371 documented incident
Privileged action without an enforceable approval gate
Limits and approvals live in a prompt, a UI checkbox or a convention instead of code the automation cannot bypass.
ASI03LLM03AC-323 documented incidents
Cross-tenant access through a caller-controlled identifier
The server trusts an account, tenant or object ID sent by the client, so changing a number reveals someone else's data.
AC-3AC-4Goal 34 documented incidents
Indirect prompt injection changes agent behavior
An email, web page, issue or ticket that the AI reads contains instructions, and the agent follows them with its tools.
ASI01ASI02ASI065 documented incidents
Destructive operation without verified recovery
A delete, cleanup or teardown runs against the wrong target, and the backups that should save the day turn out not to work.
CP-9CP-10RV.3.49 documented incidents
Retry amplification and runaway resource consumption
Retries, loops or an expensive input multiply load until a partial failure becomes a full outage or a runaway bill.
ASI08LLM06SC-58 documented incidents
Partial success leaves systems out of sync
A multi-step workflow half-completes, or events arrive out of order, and two systems now disagree about what happened.
- Confusing loan-system interface leads Citibank to wire nearly $900 million in principal to Revlon lenders (2020)
- 43-second network partition triggers database failover and unreplicated writes (GitHub, 2018)
SC-242 documented incidents
Automation acts on stale state or authorization
A background job or automation acts on a plan, permission or record that changed after it was read.
AC-3CWE-3677 documented incidents
Audit trail missing actor, target or decision
After something goes wrong nobody can tell who or what took the action, on which object, or why.
AU-3MANAGE 4.3Goal 79 documented incidents
AI action exceeds its intended scope
An AI agent or assistant takes an action, or makes a promise, that nobody meant it to be able to make.
ASI02ASI03ASI108 documented incidents
Global change without a staged rollout
A configuration, content or code change reaches every server at once, so one bad value takes everything down together.
ASI08CM-3CM-410 documented incidents
Weak or phishable authentication on a privileged path
A password, an SMS code or a help-desk reset is all that stands between an attacker and admin access.
IA-2Goal 1Goal 220 documented incidents
Known vulnerability left unremediated
A patch or fix exists, but the vulnerable component keeps running because nobody knew it was there or owned the fix.
RA-5SI-2RV.1.18 documented incidents
Insufficient segmentation lets a foothold spread
One compromised account, vendor or network segment reaches management planes, backups and everything else.
SC-7CWE-6539 documented incidents