AI Defense Matrix Catalog

Tinfoil

Visit official site ↗

Overview

Product details compiled from public sources, each with a citation.

Vendor
Tinfoil3
Description
Verifiably-private AI inference run inside hardware secure enclaves so prompts, responses, and model weights stay confidential from the provider, with an OpenAI-compatible API.4
Deployment
SaaS4
Status
Active3
Compliance
SOC 2 Type II2agent (company-level, see Methodology)

Matrix Coverage

Where this product defends, by asset class and NIST CSF function. The Coverage column shows whether each asset is Primary, Secondary, or Adjacent to what the product does. The table omits empty rows and columns.

Asset class Protect Coverage
AI Model Secondary 3
Runtime AI Data Primary 1

Framework Relevance

These frameworks include controls relevant to the asset classes Tinfoil defends. This is an editorial inference from the AI Defense Matrix asset-level crossmap, not a statement that Tinfoil implements these controls or is certified against them.

Expand Collapse
Framework Asset class Relevant controls
NIST IR 8596 AI Model Models; Algorithms (model configuration)
Runtime AI Data Prompts (runtime); inference data
CSA AI Controls Matrix AI Model Model Security; Governance, Risk and Compliance
Runtime AI Data Data Security and Privacy Lifecycle Management; Application and Interface Security
ISO 42001 AI Model A.6 AI system life cycle; A.10 Third-party and customer relationships; A.5 Assessing impacts of AI systems
Runtime AI Data A.7 Data for AI systems; A.8 Information for interested parties
Google SAIF AI Model Protect the AI model; ensure model integrity, provenance, and weight security
Runtime AI Data Expand AI red-teaming; runtime input and output safety; prompt defense
SANS Critical AI Security Guidelines AI Model Conventional Security Controls (protect model parameters with least privilege, encryption at rest, runtime obfuscation, and trusted execution environments); Data/Model Engineering Controls (adversarial training; alignment and fine-tuning); AI Supply Chain Management (public-model caution; transfer-attack exposure)
Runtime AI Data Model I/O Handling (sanitize, validate, and filter inputs and outputs; segregate user and system prompts; multilayered prompt-injection defense); Conventional Security Controls (protect augmentation and RAG data with vector-store access controls and validation); Data Minimization and Obfuscation (limit sensitive prompt content; context-window management); Limit Model Behavior (AI guardrails)
MITRE ATLAS AI Model AML.T0043 Craft Adversarial Data; AML.T0024 Exfiltration via AI Inference API (subtechniques: AML.T0024.001 Invert AI Model and AML.T0024.002 Extract AI Model); AML.T0018 Manipulate AI Model (integrity and backdoor)
Runtime AI Data AML.T0051 LLM Prompt Injection; AML.T0054 LLM Jailbreak; AML.T0056 Extract LLM System Prompt
OWASP AI Exchange AI Model Development-time and runtime model threats: model inversion, extraction, evasion, poisoning
Runtime AI Data Input threats: prompt injection, adversarial inputs, evasion; runtime threats: RAG poisoning, memory tampering
OWASP LLM Top 10 AI Model LLM03 Supply Chain; LLM04 Data and Model Poisoning; LLM09 Misinformation
Runtime AI Data LLM01 Prompt Injection; LLM02 Sensitive Information Disclosure; LLM08 Vector and Embedding Weaknesses; LLM05 Improper Output Handling
OWASP Agentic Security Top 10 AI Model ASI04 Agentic Supply Chain Vulnerabilities (model provenance, weights, and dynamic loading)
Runtime AI Data ASI06 Memory & Context Poisoning; ASI01 Agent Goal Hijack (via prompt injection in runtime inputs)

Provenance

Last sourced 2026-06-25.

Expand Collapse

Sources

  1. Tinfoil documentation introduction
    Vendor source accessed 2026-06-25
    • “User request data is processed inside secure enclaves that even Tinfoil cannot access, giving you hardware-enforced privacy that is verifiable via remote attestation.”
  2. Tinfoil trust center
    Vendor source accessed 2026-07-17
  3. Tinfoil homepage
    Vendor source accessed 2026-06-25
    • “the auditor brings the eval, the lab brings its weights, and both get cryptographic proof of exactly what ran, without the weights ever leaving the enclave.”
  4. Tinfoil inference page
    Vendor source accessed 2026-06-25

Changelog

  1. Refreshed the compliance citation against the current trust portal.

  2. Recorded the attestation type listed on the trust portal.

  3. Added to the catalog from the Tinfoil documentation.

Found an error? Corrections are welcome. Suggest an edit.