# Tamnoon Introduces Refusal-First Cloud Remediation Engine to Eliminate Production Outages Marina Segal advocates that automated cloud security must learn restraint before modifying live infrastructure. - **Source:** [The Next Web](https://brifing.si/source/afe24b20-b275-4c19-b9a6-8bb3c8babae7) - **Published:** 2026-10-13 - **Topic:** [Software & Engineering](https://brifing.si/c/engineering) - **Original article:** - **On Brifing:** https://brifing.si/s/59bea91c-f4d6-425d-aa81-082e20f1991a ## The Detection Saturation Problem Cloud security platforms successfully flag misconfigurations within minutes, but resolution remains stagnant. More than half of all detected vulnerabilities stay unresolved. While attackers achieve lateral movement in 29 minutes, enterprise remediation averages 150 days per critical finding, turning vulnerability scanners into ineffective filing systems. ## Fear of Broken Production Engineering teams resist automated remediation tools out of rational fear rather than technical reluctance. Flawed automated fixes manifest as production outages tied to human approvers. Speed is irrelevant if changes break dependencies; teams refuse to authorize machine actions that cannot prove safety beforehand. ## Architecting Autonomous Refusal Tamnoon inverted standard remediation architecture by teaching its decision engine restraint before action. Prior to altering cloud infrastructure, the engine maps resource dependencies, owner identity, traffic patterns, and historical changes. It deliberately rejects modifications when safety cannot be mathematically or contextually proven. ## Granular Multi-Verdict Processing In production tests with identical public storage alerts, the engine applies contextual triage. It automatically remediates provably safe buckets, requests human intervention for unassigned assets, and formally declines fixes that would sever live application traffic undetected by standard posture scanners. ## Evidence-Backed Audit Receipts Every refusal produces a six-field verifiable audit receipt containing underlying telemetry, specific reasoning, required reversal conditions, and manual execution steps. These receipts allow security engineers to defend operational decisions immediately, transforming arbitrary machine output into defensible compliance records for auditors and executives. ## Zero-Trust Autonomy Model Rather than relying on vendor risk toggles, Tamnoon connects via read-only permissions with autonomy set to zero. The platform earns execution rights per vulnerability class only after proving reliable analysis over time, while human engineers maintain perpetual veto authority across all cloud environments. ## Cross-Scanner Correlation Efficiency Integrating with seventeen external scanners including Wiz and CrowdStrike Falcon, the model condenses 142 noisy alerts into single actionable problems. Enterprise deployments demonstrated a reduction from 6,074 open issues to 2,041, alongside a 93 percent decline in recurring quarterly misconfigurations. ## Enterprise Cost Discrepancy Manual remediation burdens organizational budgets heavily. A healthcare enterprise customer calculated manual resolution costs at $8,775 per finding compared to $775 through verified automated workflows. Backed by $18 million in venture funding, Tamnoon counts major enterprises like Warner Music Group among clients. ## The Cloud Remediation Index Alongside a public registry of declined remediations, Tamnoon released the Cloud Remediation Index derived from 14.86 million detections. Segal predicts procurement standards will require vendors to demonstrate refusal logs, replacing synthetic detection scores with verified track records of operational restraint. ## Sources covering this story - https://thenextweb.com/news/marina-segal-tamnoon-cloud-remediation-say-no — --- _Summary written by Brifing from the original reporting. Copyright in the original article belongs to its publisher; read the full piece at the link above._