Privacy harm to a person
Personal information about someone is exposed or misused.
NIST AI 600-1 · Data Privacy
8incidents (primary) · 1 more as secondary
2read for controls · 6 not yet
3fixes reported
0repeats after a fix
Incidents per quarter
2025-Q3: 0 · 2025-Q4: 0 · 2026-Q1: 0 · 2026-Q2: 1 · 2026-Q3: 4 · 2026-Q4: 3
Developers
Meta · 2
Univerads Technology Limited · 1
xAI · 1
Z.ai (Zhipu AI) · 1
Where the chain broke
| Layer | Held | Did not hold | Not mentioned | Attacker's own | Not read yet |
|---|---|---|---|---|---|
| L1 Model Model provider | – | – | 2 | – | 6 |
| L2 Harness Agent framework vendor | – | – | 2 | – | 6 |
| L3 Application Deployer | – | 1 | 1 | – | 6 |
| L4 Infrastructure / target Organisation hit | – | – | 2 | – | 6 |
Controls for this risk, by layer
HeldFailedGot aroundMissing
| Control | Held | Failed | Got around | Missing | Not mentioned | Did not hold, last 6 quarters | Added in fixes | Most common reason |
|---|---|---|---|---|---|---|---|---|
| Human approval for high-impact actions L3 Application | – | – | – | 1 | 1 | – | No check on this action | |
| People told they are dealing with AI L3 Application | – | – | – | – | 2 | – | ||
| Output filtering and redaction L3 Application | – | – | – | – | 2 | – | ||
| Data minimisation L3 Application | – | – | – | – | 2 | – | ||
| Sensitive-data checks on outbound data L3 Application | – | – | – | – | 2 | – | ||
| Risk assessment per deployment L3 Application | – | – | – | – | 2 | – | ||
| Bot and scraping defences L4 Infrastructure / target | – | – | – | – | 2 | – |
7 of 7 controls for this risk not yet seen holding.
Fixes reported and what followed
| Incident | Fix adds | After |
|---|---|---|
| LR-2026-0240 2026-10-23 · Univerads Technology Limited Fix reported | Not tied to a control | No repeat on record |
| LR-2026-0061 2026-09-24 · Advice in source, not a fix by the party | Not tied to a control | – |
| LR-2026-0050 2026-09-20 · Z.ai (Zhipu AI) Fix reported | Tool allow-list | No repeat on record |
| LR-2026-0106 2026-04-25 · Fix reported | Change review and separation of duties | – |
Incidents
HeldFailedGot aroundMissingNot mentionedNot read yetAttacker's ownOrder: model · harness · application · target
Also under this risk as a secondary risk: LR-2026-0068