/
© 2026 RiffOn. All rights reserved.

Get your free personalized podcast brief

We scan new podcasts and send you the top 5 insights daily.

  1. The Road to Accountable AI
  2. Brett Levenson (Moonbounce): The Best AI Security System is the One People Will Use
Brett Levenson (Moonbounce): The Best AI Security System is the One People Will Use

Brett Levenson (Moonbounce): The Best AI Security System is the One People Will Use

The Road to Accountable AI · Oct 15, 2026

Moonbounce CEO Brett Levenson explains how semantic firewalls and info-first policies bring consistency and auditability to AI security.

Turn Content Moderation into Yes-No Feature Extraction to Dramatically Boost Consistency

Traditional policy review forces human evaluators to make action decisions under extreme time constraints, resulting in low agreement. Deconstructing dense policy documents into binary objective questions turns reviewers into feature extractors, drastically increasing enforcement consistency and auditability.

Brett Levenson (Moonbounce): The Best AI Security System is the One People Will Use thumbnail

Brett Levenson (Moonbounce): The Best AI Security System is the One People Will Use

The Road to Accountable AI·a month ago

Evaluate AI Action Context Instead of Relying on Static Application Permissions

Traditional cybersecurity relies on deterministic code where permissible actions are predictable. Because generative AI creates non-deterministic application layers, security controls must inspect the semantic content, intent, and context of each specific request rather than granting broad static tool access.

Brett Levenson (Moonbounce): The Best AI Security System is the One People Will Use thumbnail

Brett Levenson (Moonbounce): The Best AI Security System is the One People Will Use

The Road to Accountable AI·a month ago

Cap AI Security Interventions at Sub-300 Milliseconds to Prevent User Abandonment

Security mechanisms only work if people tolerate using them. In generative and voice AI systems, added latency damages user experience and causes abandonment, necessitating that runtime safety filtering operates within a fast, 300-millisecond broadcast-delay window.

Brett Levenson (Moonbounce): The Best AI Security System is the One People Will Use thumbnail

Brett Levenson (Moonbounce): The Best AI Security System is the One People Will Use

The Road to Accountable AI·a month ago

Prevent AI Safety Drift by Translating Judgments into Auditable Policies Rather Than Weights

Because static AI policies decay as real-world scenarios evolve, organizations need feedback loops to update rules. Converting corrections into explicit, backtested written policy rules rather than opaque model weights maintains explainability and prevents silent alignment drift over time.

Brett Levenson (Moonbounce): The Best AI Security System is the One People Will Use thumbnail

Brett Levenson (Moonbounce): The Best AI Security System is the One People Will Use

The Road to Accountable AI·a month ago

Payment Processors Function as De Facto Regulatory Enforcement for Generative Media Companies

Before statutory rules like the EU AI Act matured, merchant underwriters served as de facto regulators by threatening to cut off payment processing for companies generating questionable content. This commercial pressure forces media-generation startups to adopt fine-grained safety filters to survive.

Brett Levenson (Moonbounce): The Best AI Security System is the One People Will Use thumbnail

Brett Levenson (Moonbounce): The Best AI Security System is the One People Will Use

The Road to Accountable AI·a month ago

Replace Binary Guardrail Blocks with Real-Time Output Rewriting to Preserve User Experience

Binary allow-or-block paradigms tempt companies to disable safety controls to avoid breaking customer interactions. Introducing active transform or rewrite capabilities allows non-compliant or borderline outputs to be safely remediated on the fly without breaking application flow.

Brett Levenson (Moonbounce): The Best AI Security System is the One People Will Use thumbnail

Brett Levenson (Moonbounce): The Best AI Security System is the One People Will Use

The Road to Accountable AI·a month ago

Falsifiable and Measurable AI Standards Clarify Corporate Liability Risks

Companies deploy AI without clear safety benchmarks, fearing that extensive audit trails might create legal liabilities if flaws are recorded. Clear government standards provide explicit, measurable definitions of compliance, relieving enterprises from arbitrary legal guesswork.

Brett Levenson (Moonbounce): The Best AI Security System is the One People Will Use thumbnail

Brett Levenson (Moonbounce): The Best AI Security System is the One People Will Use

The Road to Accountable AI·a month ago