From the Anthropic edition of November 10, 2026
WIRED: researcher using consumer-grade Claude models found vulnerabilities in about two-thirds of 349 UK public-sector codebases; most were basic hygiene failures
WIRED (reporting on ControlPlane research) · Technology press · United Kingdom
- Publisher
- WIRED
- Original publication
- November 10, 2026
- Source retrieved
- 2026-11-11
What happened
WIRED reports that ControlPlane researcher Chris Nesbitt-Smith, using consumer-grade Claude Opus and Sonnet models on a laptop, found vulnerabilities in about two-thirds of 349 UK public-sector codebases across 33 organizations, including a Planning Inspectorate credential committed to GitHub that would have given broad SharePoint and OneDrive access. The Planning Inspectorate said its monitoring detected the activity, it fixed the issue within 20 minutes of the report and found no breach. The researcher says many issues could have been caught by a routine IT health check and warns that closing public repositories without fixing flaws is the worst outcome.
Why it matters
An independent account of how cheaply Claude-class models can audit public-sector code, and of agency responses. U.S. state, local and education agencies that publish code, or use contractors who do, should expect AI-assisted researchers and attackers to find credential leaks and access-control gaps, and should prepare disclosure handling rather than retreat from open code.
Evidence and measured results
Credible investigative reporting with agency responses and an independent industry comment (OpenUK); quantitative claims originate from the researcher.
Limitations and uncertainty
Journalistic account relying on the researcher and agency statements; numbers not independently verified; does not detail AI false positives. Cross-tagged state-government and local-government: report covers UK departments and notes local authorities publish code; lessons transfer to U.S. agencies publishing code.
Put this evidence to work
Lighthouse Advisory interpretation, grounded in this source. Enriched 2026-11-11; this does not change the original publication date. Each role follows the same checklist so takeaways compare across sources.
Sales
Role takeaway
- Customer problem
- Agencies publishing code have exposed secrets and access-control gaps that AI-assisted researchers can now find cheaply.
- Stakeholders
- Agency CISO
- Identity and access management lead
- Open-source program office
- Vendor management
- Discovery questions
- When did you last scan repository history for secrets?
- How are app registration permissions scoped?
- Do you have a vulnerability disclosure policy?
- Would you detect use of a leaked credential?
- Value hypothesis
- A hygiene assessment could close common, high-impact exposures before outside researchers or attackers find them.
- Potential engagement
- A repository and credential hygiene assessment with a disclosure runbook.
- Unsupported claims
- That Claude enables novel attacks
- That closing repositories removes exposure
- Independent verification of all counts
Pre-sales engineering
Role takeaway
- Fit
- Applies to any agency or university with public code.
- Architecture and integration
- Secret scanning in CI; identity provider rotation; access-control tests.
- Prerequisites
- Repository inventory
- Identity provider admin access
- Constraints
- Researcher-reported numbers
- UK context
- Security
- Least-privilege credentials, rotation on exposure, monitoring for misuse.
- Proof of value
- Run history scan on all public repos
- Test detection of a canary credential
- Verify access checks on request-ID endpoints
Delivery
Role takeaway
- Work
- Audit repos and credentials and prepare disclosure response.
- Dependencies
- Identity team
- Vendors
- Code owners
- Ownership
- Agency security office.
- Skills and adoption
- Developers trained on secret handling.
- Governance checkpoints
- Disclosure policy published
- Credential rotation verified
- Quarterly scan review
- Acceptance criteria
- Proposed: zero valid secrets in public repo history after remediation
- Proposed: canary credential use detected within one hour
- Risks
- Long-lived broad credentials
- Slow vendor fixes
- Closing repos without fixes
Implementation considerations
Lighthouse Advisory interpretation across the operating dimensions a public-sector buyer must settle before this evidence becomes a design. Each note answers the question under its heading for this specific source.
Architecture and integration
What must connect, and where does the AI sit in the workflow?
Credential hygiene (short-lived, least-privilege secrets; history scanning) matters more than the AI tool; deleting a secret from a branch does not revoke it.
Governance
Who approves, reviews and stays accountable for outcomes?
Policy debates about withdrawing open code (the NHS reportedly ordered a move away in May) should weigh the researcher's warning that mirrored repositories remain exposed.
Security and privacy
What data, permissions and controls need testing?
Exposed Entra ID credential with broad permissions; access-control gaps in Department for Education services; Scottish Government portal flaw exposing administrative account information; all dealt with per WIRED.
Accessibility and workforce
Who is affected, and what skills or accommodations follow?
Not addressed in the source.
Procurement
What should contracts, pricing and exit terms secure?
Contracts with software vendors should require secret scanning and vulnerability disclosure handling for code published on the agency's behalf.
Operating model
Which teams own the service once it runs?
Agency monitoring that detects researcher activity (as the Planning Inspectorate reports) is part of the response model.
Publication history
- 2026-11-10Anthropic · Issue 1210 resources
Stable resource ID: wired-uk-government-code-flaws-claude-researcher-planning-inspectorate-20261110