Last updated: 1 August 2026.

Current AI inventory and roles

CertentiScan can generate technical remediation guidance using the commercial Anthropic Claude API. Anthropic provides the underlying model. Burhuc deploys Claude and, as a conservative compliance position, also treats itself as the provider of the branded AI-fix feature. The feature does not make decisions about people or determine access to employment, credit, education, insurance, public services, or other regulated opportunities.

Transparency and provenance

The scanner identifies the feature before use as AI-generated and names Claude as the provider. Generated guidance carries visible origin information, machine-readable provider and model metadata, creation time, and a durable provenance marker when copied. Users are told to review the output before use.

Article 50 transparency status

From 2 August 2026, the AI Act transparency rules for direct AI interaction, generated-content marking, deepfake disclosure, and certain AI-generated public-interest text apply. Burhuc treats this as an operating control: user-facing AI features must identify AI involvement before meaningful use, AI-generated technical guidance must carry provider and model provenance where appropriate, and public publications must keep visible human-review and editorial-responsibility statements.

Publications and human editorial review

Siggi is Burhuc's editorial agent persona. AI may assist with drafting, but every Burhuc publication is reviewed and approved by a person before it is published. Burhuc Research holds editorial responsibility for the final text. Publication pages carry a visible human-review statement and machine-readable editorial-review metadata.

AI literacy and operating controls

People operating Burhuc AI systems or reviewing AI-assisted content receive role-specific guidance covering system limits, hallucinations, personal-data handling, human review, transparency, escalation, and prohibited or high-risk uses. The inventory and guidance are reviewed when a model, purpose, data flow, or public feature changes and at least annually.

Risk boundary

The current public AI-fix feature is a technical assistance tool. It is not presented as a high-risk AI system or a prohibited AI practice. A new AI use must be classified before release and cannot be used for consequential decisions without a dedicated legal and technical assessment.

Related sources

For official EU information, see the European Commission's Article 4 AI-literacy text, Article 50 transparency text, and Code of Practice on transparency of AI-generated content. The Commission also publishes its 2 August 2026 enforcement notice, Article 50 transparency FAQs, and AI Act enforcement information.

This page describes Burhuc's operational posture. It is informational and is not a regulatory certification or legal opinion.

Contact

Questions about Burhuc AI governance can be sent to contact@burhuc.com.