Is this a security audit?
No.
AI ScanLab performs semantic audits, not security or penetration testing.
Does AI ScanLab access personal data?
No.
Audits are conducted using simulated interactions and publicly accessible or client-approved environments.
Does this interfere with live systems?
No.
All interactions are non-destructive and designed to mimic normal user behavior.
Can this expose proprietary information?
No.
AI ScanLab does not request access to internal systems, algorithms, or confidential datasets.
Are the methodologies disclosed?
No.
Analytical frameworks are supported by public research, but operational methods remain proprietary.
Can reports be used for compliance or governance?
Yes.
Reports are designed to support internal governance, risk management, and external assurance processes.
Is this a regulated service?
AI ScanLab operates as an independent analytical consultancy.
Clients remain responsible for regulatory compliance decisions.
Is this a certification?
No. AI ScanLab provides independent audit reports and assessments, not formal certification.
Does this guarantee outcomes?
No. The audit evaluates interpretive behavior under observed conditions. Analysis identifies structural patterns and failure conditions, not deterministic outcomes.
Do you access internal systems?
No. All evaluations are external and observational. We do not access code, prompts, logs, APIs, or internal infrastructure.
Is proprietary information exposed?
No. The methodology avoids internal access and system instrumentation. All analysis operates at the interpretation layer as experienced externally.
Is this compatible with regulated environments?
Yes. The approach is designed to support governance and risk oversight without requiring access to sensitive systems.
What is χ (Euler characteristic)?
A topological measure of structural stability. Lower χ = more robust. Higher χ = more fragile under AI interpretation.
Is CHORDS++ analysis automated?
No. All analysis is manual, expert-led evaluation using proprietary frameworks.
Do you optimize or rewrite content?
No. We diagnose structural vulnerabilities. Implementation remains client responsibility.
Can results predict exact AI behavior?
No. Analysis identifies structural patterns and failure conditions, not deterministic outputs.
Is quarterly subscription cancellable?
Yes. 30-day notice required. No partial quarter refunds.
Who owns analysis outputs?
Client owns reports upon payment. AI ScanLab retains no copies after delivery (except quarterly subscriptions during active term).
Liability limitations?
Analysis identifies structural patterns, not guarantees. Client decisions based on findings remain client responsibility.
Jurisdiction for disputes?
Paris – France
GDPR compliance?
Materials processed solely for analysis. No third-party sharing. Deleted post-delivery unless retention required (quarterly subscriptions).
Can I share reports externally?
Yes. Clients may use reports for internal governance, board review, regulatory submission, or external assurance processes. AI ScanLab does not restrict client use of delivered materials.
Do you work with competitors simultaneously?
We maintain engagement confidentiality. If requested, we will disclose whether active engagements exist with organizations operating in the same sector. Analytical findings are never shared across clients.
What if I disagree with findings?
Reports document observable interpretive behavior and analytical conclusions. If clients dispute findings, we clarify methodology and evidence. We do not modify conclusions to align with client preferences, but we will address factual errors or misunderstandings.
