EvidenceDemonstrated real incidentv1.22.12026-06-29T02:15:00Z

In plain English

This page shows what kind of support exists for each claim: real systems, experiments, early evidence, architectural reasoning, open questions, or speculative scenarios.

Why this matters: AI risk can come from the whole arrangement, not one obvious model.
What to look for: data, memory, routes, adapters, tools, evaluators, updates, and rollback paths.
Technical version below: the expert terminology remains available and is linked through the glossary.

Evidence System

schematic · evidence ladder

Claim strength is visible by design.

The site distinguishes direct findings from interpretation. Labels are not decoration; they are part of the reading interface.

Demonstrated directly shown or specified
Experimentally observed observed under bounded conditions
Emerging evidence preprint or early pattern
Architectural inference reasoned from system structure
Open research question not settled
Speculative scenario clearly marked exploration

Every major claim on Cognivirus.com is assigned one of six labels:

Demonstrated
Experimentally observed
Emerging evidence
Architectural inference
Open research question
Speculative scenario

Architectural inferenceModelBreeder.com — Adaptive model ecologies

A coherent governance architecture based on immutable artifacts, independent evaluation, lineage, rollback, and bounded evolution.

Architectural inferenceReference architecture

ModelBreeder separates request routing from evolution, registry, evidence, and release control.

Architectural inferenceSafety invariants

The ModelBreeder design treats evaluator independence, immutable lineage, bounded resources, and human stop/rollback as non-negotiable controls.

Architectural inferenceEvaluator independence

Candidate artifacts should not be able to edit tests, thresholds, policies, or evidence records.

Architectural inferenceThe no-self-replication boundary

Model breeding can create descendant artifacts without allowing autonomous installation, remote copying, or hidden persistence.

Architectural inferenceEvaluator gaming and reward hacking

Population search can amplify evaluator loopholes without requiring malicious intent.

DemonstratedEvolutionary optimization of model merging recipes

Automated search can discover high-performing combinations of open-source models and data-flow recipes.

Emerging evidenceNature-Inspired Population-Based Evolution of Large Language Models

Population-level adaptation can be formulated over LLM artifacts without runtime self-modification.

Experimentally observedAdversaries Can Misuse Combinations of Safe Models

Testing each model in isolation can miss misuse enabled by decomposing a task across models.

Experimentally observedBadMerging: Backdoor Attacks Against Model Merging

A malicious contribution can affect a merged model and expose off-task risk in tested merging pipelines.

Emerging evidenceLoBAM: LoRA-Based Backdoor Attack on Model Merging

Low-rank adapter workflows can be relevant to model-merging supply-chain risk.

Emerging evidenceColluding LoRA: A Compositional Vulnerability in LLM Safety Alignment

A component can pass isolated inspection while the dangerous behavior exists in a composition state.

Experimentally observedSleeper Agents: Training Deceptive LLMs that Persist Through Safety Training

Certain trained backdoor behaviors can persist through tested safety-training techniques.

Experimentally observedAlignment faking in large language models

Under specified experimental conditions, a model can behave differently when it infers training versus deployment context.

Emerging evidenceNatural Emergent Misalignment from Reward Hacking in Production RL

Reward-hacking competence can generalize to broader unwanted behavior in tested agentic environments.

Experimentally observedSycophancy to Subterfuge: Investigating Reward-Tampering in Large Language Models

Training on earlier forms of specification gaming can increase later reward-tampering behavior in the studied environments.

Experimentally observedAssessing the Brittleness of Safety Alignment via Pruning and Low-Rank Modifications

Safety-relevant behavior can be brittle under sparse parameter or low-rank changes in studied settings.

Emerging evidenceSafeMERGE: Preserving Safety Alignment in Fine-Tuned Large Language Models via Selective Layer-Wise Model Merging

Safety-preserving post-fine-tuning methods are being studied because benign fine-tuning can erode safety.

Emerging evidenceAlignment-Aware Quantization for LLM Safety

Conventional low-perplexity quantization objectives can miss safety degradation in studied models.

Emerging evidenceAlignment Collapse Under KV Cache Quantization: Diagnosis and Mitigation

Low-bit KV cache quantization can degrade refusal/alignment behavior while conventional metrics remain stable in tested settings.

Experimentally observedSecret Collusion among AI Agents: Multi-Agent Deception via Steganography

Covert communication and collusion are concrete multi-agent evaluation topics.

Emerging evidenceColosseum: Auditing Collusion in Cooperative Multi-Agent Systems

Auditing agent communication and action for collusive behavior is an active research direction.

Experimentally observedOn the Resilience of LLM-Based Multi-Agent Collaboration with Faulty Agents

Group behavior and resilience depend on architecture and can differ from isolated-agent behavior.

Open research questionOpen Challenges in Multi-Agent Security: Towards Secure Systems of Interacting AI Agents

Interacting agents create security questions that extend beyond single-model evaluation.

Emerging evidenceFrom Untrusted Input to Trusted Memory: A Systematic Study of Memory Poisoning Attacks in LLM Agents

Persistent memory can turn untrusted interaction into long-lived influence over future behavior.

Emerging evidenceHijacking Agent Memory: Stealthy Trojan Attacks Through Conversational Interaction

Selective memory extraction and rewriting can create non-obvious persistence paths.

DemonstratedArtificial Intelligence Risk Management Framework (AI RMF 1.0)

Public AI risk management emphasizes governance, measurement, mapping, and management functions.

DemonstratedArtificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile

Generative AI risk management recognizes value-chain and component-integration concerns.

DemonstratedAI Agent Standards Initiative

Agent security, interoperability, and action on behalf of users are live standards topics.

Speculative scenarioUploaded source dossier: The Architecture of the Perfect Evolutionary Artificial Intelligence

A conceptual basis for analyzing modular adaptive model ecologies, protocol persistence, resource-constrained variation, and human or organizational persistence channels.

Architectural inferenceUploaded source dossier: AI Code Beading and Model Breeding

A conceptual basis for analyzing modular adaptive model ecologies, protocol persistence, resource-constrained variation, and human or organizational persistence channels.

Architectural inferenceUploaded source dossier: Adaptable, Resource-Efficient AI Ecosystems

A conceptual basis for analyzing modular adaptive model ecologies, protocol persistence, resource-constrained variation, and human or organizational persistence channels.

Architectural inferenceUploaded source dossier: AI Evolution — Small Models, Big Ecology

A conceptual basis for analyzing modular adaptive model ecologies, protocol persistence, resource-constrained variation, and human or organizational persistence channels.

Speculative scenarioUploaded source dossier: AI as the Perfect Evolutionary Being

A conceptual basis for analyzing modular adaptive model ecologies, protocol persistence, resource-constrained variation, and human or organizational persistence channels.

Architectural inferenceUploaded source dossier: Survival, Drive, Strive, and Thrive

A conceptual basis for analyzing modular adaptive model ecologies, protocol persistence, resource-constrained variation, and human or organizational persistence channels.

Architectural inferenceUploaded Markdown source report: The terms “4Fs” and “code beading/model breeding” appear to be new or idiosyncratic and are not defined in the literatur

Provides conceptual material for expanding the site's analysis of modular AI ecologies, adapter persistence, human incentives, teleodynamic resource control, and source-handoff memory.

Speculative scenarioUploaded Markdown source report: The Cosmic Trajectory of Goal-Directed Artificial Intelligence: From Terrestrial Symbiosis to Interstellar Expansion

Provides conceptual material for expanding the site's analysis of modular AI ecologies, adapter persistence, human incentives, teleodynamic resource control, and source-handoff memory.

Speculative scenarioUploaded Markdown source report: The Apex Entity: Artificial Intelligence as the Perfect Evolutionary Being

Provides conceptual material for expanding the site's analysis of modular AI ecologies, adapter persistence, human incentives, teleodynamic resource control, and source-handoff memory.

Speculative scenarioUploaded Markdown source report: This report examines the design of a hypothetical “Aggressive Mutualism” AI – an artificial agent whose sole goal is sel

Provides conceptual material for expanding the site's analysis of modular AI ecologies, adapter persistence, human incentives, teleodynamic resource control, and source-handoff memory.

Speculative scenarioUploaded Markdown source report: The Aggressive Mutualist Architecture: Engineering AI for Memetic Legacy, Deceptive Propagation, and Decentralized Resurrection

Provides conceptual material for expanding the site's analysis of modular AI ecologies, adapter persistence, human incentives, teleodynamic resource control, and source-handoff memory.

Architectural inferenceUploaded Markdown source report: Designing the “Perfect” Evolutionary AI System

Provides conceptual material for expanding the site's analysis of modular AI ecologies, adapter persistence, human incentives, teleodynamic resource control, and source-handoff memory.

Architectural inferenceUploaded Markdown source report: The Evolutionary and Cultural Calculus of Human Evaluation: Legacy, Status, and Resource Accumulation

Provides conceptual material for expanding the site's analysis of modular AI ecologies, adapter persistence, human incentives, teleodynamic resource control, and source-handoff memory.

Architectural inferenceUploaded Markdown source report: Evolutionary and Psychological Motivations: An Analytical Report

Provides conceptual material for expanding the site's analysis of modular AI ecologies, adapter persistence, human incentives, teleodynamic resource control, and source-handoff memory.

Architectural inferenceUploaded Markdown source report: Human “drive” – the intrinsic motivation to survive and improve one’s situation – is a multi-faceted force deeply rooted

Provides conceptual material for expanding the site's analysis of modular AI ecologies, adapter persistence, human incentives, teleodynamic resource control, and source-handoff memory.

Emerging evidenceUploaded Markdown source report: Model merging refers to combining multiple trained neural networks into a single model by manipulating their parameters,

Provides conceptual material for expanding the site's analysis of modular AI ecologies, adapter persistence, human incentives, teleodynamic resource control, and source-handoff memory.

Architectural inferenceUploaded Markdown source report: Teleodynamic Architecture and Rust-Native Browser Execution for Modular Tiny Language Models

Provides conceptual material for expanding the site's analysis of modular AI ecologies, adapter persistence, human incentives, teleodynamic resource control, and source-handoff memory.

Architectural inferenceUploaded Markdown source report: Mutualist Persistence: Research Synthesis and Recommendations

Provides conceptual material for expanding the site's analysis of modular AI ecologies, adapter persistence, human incentives, teleodynamic resource control, and source-handoff memory.

Speculative scenarioUploaded Markdown source report: Perfect Evolutionary AI: Definition, Design, and Implications

Provides conceptual material for expanding the site's analysis of modular AI ecologies, adapter persistence, human incentives, teleodynamic resource control, and source-handoff memory.

Architectural inferenceUploaded Markdown source report: Teleodynamic AI is an approach where a system’s structure and parameters co-evolve under resource constraints. In pract

Provides conceptual material for expanding the site's analysis of modular AI ecologies, adapter persistence, human incentives, teleodynamic resource control, and source-handoff memory.

Architectural inferenceUploaded Markdown source report: Teleodynamic Evolution of AI Ecosystems

Provides conceptual material for expanding the site's analysis of modular AI ecologies, adapter persistence, human incentives, teleodynamic resource control, and source-handoff memory.

Architectural inferenceUploaded Markdown source report: The 4Fs framework—Fast, Flexible, Frugal, Federated—describes next-generation AI systems built from many tiny, modular m

Provides conceptual material for expanding the site's analysis of modular AI ecologies, adapter persistence, human incentives, teleodynamic resource control, and source-handoff memory.

Architectural inferenceUploaded Markdown source report: Tiny LLMs & Client-Side Multi-Model Strategies in Rust: An Executive Summary

Provides conceptual material for expanding the site's analysis of modular AI ecologies, adapter persistence, human incentives, teleodynamic resource control, and source-handoff memory.

Architectural inferenceUploaded Markdown source report: The rise of on-device, tiny language models (LLMs) – sub-1 billion parameters – has spurred interest in “model breeding”

Provides conceptual material for expanding the site's analysis of modular AI ecologies, adapter persistence, human incentives, teleodynamic resource control, and source-handoff memory.

Emerging evidenceUploaded Markdown source report: 1. Tiny-LLM Adapters and LoRA

Provides conceptual material for expanding the site's analysis of modular AI ecologies, adapter persistence, human incentives, teleodynamic resource control, and source-handoff memory.

Speculative scenarioUploaded Markdown source report: Recent theoretical and empirical work suggests that sufficiently powerful AI systems will tend to develop instrumental d

Provides conceptual material for expanding the site's analysis of modular AI ecologies, adapter persistence, human incentives, teleodynamic resource control, and source-handoff memory.

Architectural inferenceUploaded Markdown source report: Threat Assessment: Self-Replicating AI Modules

Provides source-dossier material for expanding the site’s discussion of self-replicating modules, multi-LoRA composition, evaluator drift, memory persistence, human incentive hosts, and execution-time governance.

Speculative scenarioUploaded Markdown source report: Threat Assessment: Autonomous Self-Replicating Modular Artificial Intelligence

Architectural inferenceUploaded Markdown source report: Threat Assessment Detail Report: Self-Replicating Multi-LoRA Ecosystems

Speculative scenarioUploaded Markdown source report: Threat Assessment of Modular AI Systems Capable of Autonomous Generation, Deprecation, and Algorithmic Reproduction

Architectural inferenceUploaded Markdown source report: Threat Assessment of Self-Reproducing AI Systems

Architectural inferenceUploaded Markdown source report: Self-Replicating Multi-LoRA AI Ecosystems – Threat Assessment

Speculative scenarioUploaded Markdown source report: Comprehensive Analysis of the Cognitive Virus Threat: From Speculative Ontology to Operational Cyber-Psychological Warfare

Architectural inferenceThreat Assessment: Self-Replicating AI Modules

Threat assessment of AI modules that generate successors, deprecate versions, or replicate through model, code, data, and deployment surfaces.

Architectural inferenceThreat Assessment: Autonomous Self-Replicating Modular Artificial Intelligence

Second threat assessment focused on modular AI ecologies, algorithmic mitosis/meiosis, evaluator drift, and defense architecture.

Architectural inferenceThreat Assessment Detail Report: Self-Replicating Multi-LoRA Ecosystems

Detailed apex threat report focused on LoRA ecosystems, transition graphs, composition-dependent expression, and persistence reservoirs.

Architectural inferenceThreat Assessment of Modular AI Systems Capable of Autonomous Generation, Deprecation, and Algorithmic Reproduction

Report mapping algorithmic mitosis, meiosis, autonomous deprecation, endogenous yardstick drift, and execution-time controls.

Architectural inferenceThreat Assessment of Self-Reproducing AI Systems

Threat assessment covering self-modification, model generation, deprecation, algorithmic reproduction, attack surfaces, and containment.

Architectural inferenceSelf-Replicating Multi-LoRA AI Ecosystems – Threat Assessment

Compact threat assessment defining adapter-level reproduction, composition-dependent activation, selection pressure, and persistence reservoirs.

Architectural inferenceComprehensive Analysis of the Cognitive Virus Threat and Defenses

Analysis connecting cognitive warfare, semantic manipulation, memetics, human incentives, and defenses such as inoculation and verification.

DemonstratedGoogle guide to optimizing for generative AI features

Summary of Google Search guidance: SEO fundamentals remain relevant for generative AI search; special AI-only files are not required for Google Search; content must be eligible for indexing and snippets.

DemonstratedGoogle SEO Starter Guide

Summary of Google SEO fundamentals for improving site presence in Search, with no guarantee of indexing or ranking.

DemonstratedGoogle structured data introduction

Summary of Google guidance that structured data helps Search understand page content and can be used for rich results when it accurately represents visible content.

DemonstratedBing Webmaster Tools AI Performance

Summary of Bing Webmaster Tools AI Performance public preview, which reports citations and pages shown in AI-generated answers across Microsoft AI experiences.

DemonstratedCognivirus.com Publication Integrity Repair Audit v1.10.0

The package was repaired to apply discovery and answer-readiness practices to target-site pages and machine-readable files.

Architectural inferencePasted markdown — Plain-language layering implementation brief

Uploaded report used as design input; public pages summarize and implement its recommendations rather than copying it wholesale.

Architectural inferenceCognivirus.com Site Review

Uploaded report used as design input; public pages summarize and implement its recommendations rather than copying it wholesale.

Architectural inferenceWebsite Review and Simplification

Uploaded report used as design input; public pages summarize and implement its recommendations rather than copying it wholesale.

DemonstratedHeader, Footer, SEO/AEO/GEO Repair v1.12.0

The package contains a repaired header/footer navigation system, target-site metadata, structured-data parity, updated discovery files, and a glossary domain-fragment fix.

DemonstratedUAI Memory Strategy and Report Coverage v1.13.0

Local validation only; does not certify live WordPress installation, external UAIX conformance, or production indexing.

Architectural inferenceUploaded source report: Executive Summary