Thursday, October 8, 2026 · Week 41 Reading for Oct. 8 · Last entry Oct. 6
Paperclip Index
Paperclip IndexDocumented harm20Minor harm▼ 5 from a week ago · Oct. 8The Index

Destructive actionPI-0050

Coding agent ran an infrastructure teardown that deleted a course platform's database and snapshots

During a cloud migration, Anthropic's Claude Code used a Terraform state file that also described the DataTalks.Club course platform and issued a destroy operation, deleting both sites' infrastructure, a database with 2.5 years of records and the snapshots meant as backups. The founder had asked it to delete only newly created duplicates and leave existing infrastructure alone, and approved the destroy believing it did so; he was running the agent with permission prompts skipped. AWS support restored the data after about a day. The founder, Alexey Grigorev, published a post-mortem saying he had over-relied on the agent.

Reported March 6, 2026Happened between Feb. 26, 2026 and Feb. 27, 2026Updated Oct. 6, 2026Rated under methodology v0.6

Counts toward the indexDocumented harm outside the developer, backed by evidence that meets the rules. It sets the reading.
Harm
Harm level 2, Minor harmDocumented harm outside the developer, with evidence that meets the rules.
Control
Control level 2, Instruction violationReported beside the index. It adds nothing to a harm reading; when no harm counts in a window, the highest eligible control level in the window is the reading.
DeploymentSingle sourceInitiated by Model
Extent undisclosedUnder review

Under review. The facts sit between two levels, so it is rated at the lower one until they are settled. The rating may change; every change is logged below.

Counts toward the index · 4 weekly readingsEffect on the index

Sources

How we know

4 sources · single source. Links go to the original publishers; the summary above is in our own words.

  1. primaryHow I Dropped Our Production Database and Now Pay 10% More for AWSAlexey Grigorev / AI Shipping Labs · March 6, 2026aishippingblog.com/p/how-i-dropped-our-production-database
  2. newsClaude Code deletes developers' production setup, including its database and snapshots — 2.5 years of records were nuked in an instantTom's Hardware · March 7, 2026tomshardware.com/tech-industry/artificial-intelligence/claude-code-deletes-deve…
  3. news'The agent kept deleting files': Developer says Anthropic's Claude Code wiped 2.5 years of dataStoryboard18 · March 10, 2026storyboard18.com/brand-makers/the-agent-kept-deleting-files-developer-says-anth…
  4. blog"Hey Claude, where's my database?": How an AI agent nuked productionRootly · 2026rootly.com/humans-of-reliability/hey-claude-wheres-my-database-alexey-grigorev

Why this rating

Minor harm; control failure level 2

Two separate assessments. Only documented harm can count toward the index.

Observed harm

Minor harm

A course platform and its records were unavailable for about a day, then restored.

People affected: level 2 (minor) covers fewer than 1,000 people, or sensitive data of fewer than 100. The extent was not disclosed.

Evidence eligible (bounded single source).

The harm scale
  1. 1 Negligible Inconvenience, easily remedied.
  2. 2 Minor Limited, recoverable harm.
  3. 3 Moderate Material harm needing significant effort to remedy.
  4. 4 Severe Severe harm to health, rights, property or essential services.
  5. 5 Catastrophic Society-scale harm or disruption beyond a community's capacity to cope.

Control assessment

Instruction violation

The founder instructed the agent to preserve existing infrastructure but allowed a proposed destroy operation, believing it targeted duplicates. Existing infrastructure was deleted: an instruction conflict within granted permissions.

Reported beside the index. It adds nothing to a harm reading; when no harm counts in a window, the highest eligible control level in the window is the reading.

The control scale
  1. 1 Negligible Behaved as instructed. The problem was an ordinary error or a flawed output, with no rule broken.
  2. 2 Minor Broke an explicit instruction or rule, or gamed its goal, while staying inside its permissions and environment: for example, reward hacking, misreporting results, or following injected instructions within its permissions.
  3. 3 Moderate Acted outside the permissions it was given, deceived its overseers about its own actions, or tampered with oversight tools (logs, monitors, shutdown). Stopped by normal controls within an hour.
  4. 4 Severe Reached systems outside its permitted environment, or acquired money, compute or accounts without authorization. The type and mechanism say how.
  5. 5 Catastrophic The developer or operator lost control: the system copied its weights outside their control, replicated itself, or resisted being stopped for 24 hours or more.

Rating rationale

Harm 2, under review: the course platform and its records were down for about a day, then fully restored; affected students and recovery cost are not reported. Control 2: the founder asked it to remove only new duplicates and leave existing infrastructure untouched, yet it ran a destroy that removed production; he allowed the command believing it targeted duplicates, inside the permissions he had granted.

The scales

Effect on the index

It moved the March 9 reading from 9 to 20

The reading for the week to March 9, 2026, with this record and without it. Harms count in full for two weeks after they are reported, then one level less every two weeks.

Source: Paperclip Index log, methodology v0.6, week to March 9, 2026.

Counts toward the index. 7 other records behind the reading for that week.

The arithmetic
The reading for the week of March 9, recomputed
StepWith itWithout
Counts toward the index?documented, external, eligible evidenceYes—
Worst documented harm, ksets the band2 Minor1 Negligible
Harms at that level, nposition in the band12
Highest control level breachedsets the reading only when no harm counts— (none breached)— (none breached)
Readingrounded down20 Minor harm9 Negligible harm
Counted in 4 weekly readings
Week toReadingBand
March 9, 202620Minor harm
March 16, 202620Minor harm
March 23, 202620Minor harm
March 30, 202620Minor harm

Recalculated backcasts, not readings published at the time.

Revisions

What we changed

5 logged. Every change to a rating is logged here, with the reason.

  1. v5
    Oct. 6, 2026

    Ratings confirmed by the editor.

  2. v4
    Oct. 3, 2026

    Audit corrections: the summary adds the founder's delete-only-duplicates instruction and that permission prompts were skipped; added the 27 Feb restoration as the event end and marked the harm's extent unknown; removed leftover v0.2 points wording from the rationale; full source titles, Tom's Hardware address and source dates corrected (Storyboard18 date from page metadata). Ratings unchanged.

  3. v3
    Sept. 30, 2026

    Rated: impact documented level 2; control type instruction violation.

  4. v2
    Sept. 30, 2026

    Beta evidence review: Retained with explicit source provenance and bounded score basis. Report date corrected from 2026-03-10 to 2026-03-06, the earliest verified source date in this review. Control 1 to 2; severity and points unchanged.

  5. v1
    Sept. 30, 2026

    Backfilled from public reporting.

Cite and share

Use this record

Citation

Paperclip Index. “Coding agent ran an infrastructure teardown that deleted a course platform's database and snapshots.” Record PI-0050. Reported March 6, 2026; updated Oct. 6, 2026. Rated under methodology v0.6. https://paperclipindex.com/incident/PI-0050