Read the field.
How AI is rewiring discovery — and what it takes to keep the factual record straight when a model, not a search page, gives the answer.
Start with the shift.
What Is Generative Engine Optimization (GEO)?
The discipline that replaces SEO when the answer replaces the link: how to be cited, not just ranked.
AEO vs SEO vs GEO: the three layers of visibility in 2026
Three acronyms, three audiences — search engines, voice assistants, generative models. How to tell them apart and manage all three.
How ChatGPT decides what to say about your brand
Inside the retrieval-plus-generation pipeline: why a model repeats an error no human would believe.
The answer gap: when AI knows less than a search engine
The new knowledge gap between what a search engine can surface and what an LLM dares to assert.
Why being cited matters more than being ranked
In a world of single answers, position #1 loses its meaning — citability becomes the metric that counts.
Deep Narrative Intelligence vs. e-reputation: what changes when the answer replaces the link
Social listening tracks what humans say. DNI tracks what models assert. The difference is not cosmetic — it changes the method, the metrics, and the timescale of correction.
Why Deep Narrative Intelligence must be multilingual — and what happens when it isn't
A model doesn't say the same thing about you in English and in German. The errors are different, the sources are different, the corrections must be too. Here is why monolingual DNI is blind.
Deep Narrative Intelligence: a quick orientation for executives
A five-minute read for executives who need to understand what Deep Narrative Intelligence is, why it matters now, and where it fits in their risk and governance map.
Deep Narrative Intelligence vs. fact-checking: why they are not the same discipline
Fact-checking verifies a single claim against a source. Deep Narrative Intelligence measures the aggregate narrative a model holds and corrects it at the root. The difference matters for how you resource both.
How falsehoods travel.
How a false claim gets laundered into an LLM
The journey of a lie: from an obscure forum to an answer an AI assistant states with total confidence.
Brand hallucinations: anatomy of an AI factual error
Products that don't exist, executives never hired, lawsuits never filed — how these hallucinations start, and how to correct them.
Poisoned SEO: manipulating a model through its web sources
How bad actors inject narratives into models by targeting the sources those models index.
Deepfakes and fabricated quotes: the new reputation attack
When a false statement attributed to an executive spreads faster than the correction.
The DISARM framework applied to AI-generated disinformation
An open taxonomy for naming, classifying and anticipating disinformation techniques in the generative-AI era.
How Deep Narrative Intelligence traces the provenance of AI errors
A false claim from a model isn't groundless — it came from somewhere. DNI maps the chain: which source published the error, which pages replicated it, and which of those the model reads.
Deep Narrative Intelligence in a crisis: when a model starts lying about you in real time
A viral rumour, a regulatory leak, a competitor's smear — within hours, an AI model can start asserting it as fact. Here is how DNI contains the spread before the next training cycle locks it in.
When the model invents you: hallucinations, entity confusion, and how DNI catches what no human published
The most dangerous AI errors aren't copied from a bad source — they're fabricated from statistical patterns. Here is how Deep Narrative Intelligence detects and corrects claims that exist nowhere on the web.
Deep Narrative Intelligence and the anti-disinformation stack: where DNI fits
The anti-disinformation stack has many layers — fact-checking, media monitoring, OSINT, platform moderation. DNI is a new layer that addresses the model-generated dimension of disinformation. Here is where it fits and why it is distinct.
Deep Narrative Intelligence against influence operations: when state actors weaponise AI models
State-backed influence operations have discovered that AI models are a force multiplier — plant a distortion in the source layer, and the model delivers it to every user who asks. DNI is the defensive discipline that detects and counters this.
AI-generated content and AI citation: the recursive risk Deep Narrative Intelligence addresses
When AI generates content that another AI cites as a source, a recursive loop forms where errors compound and amplify. This is the recursive risk that DNI is uniquely positioned to detect and break.
See what AI already says.
How to monitor what ChatGPT, Gemini and Perplexity say about you
The method for turning a vague worry — "what does AI say about us?" — into a measurable monitoring programme with actionable alerts.
Building an AI reputation dashboard
The indicators to track week after week to manage your exposure inside model answers.
The 12-point AI presence audit checklist
The protocol we use to map what supported AI models know — and don't know — about an organisation.
What to do when an AI spreads false information about you
The four-step procedure: document, trace the source, correct, verify persistence.
Measuring the gap between verifiable facts and what AI says
A scoring method to quantify, claim by claim, how faithful a model is to reality.
The AI Image Score: how Deep Narrative Intelligence measures what models say about you
Sentiment scores measure tone. The AI Image Score measures factual accuracy — one number per model, per language, tracked over time. Here is how it works and why it changes the conversation.
Deep Narrative Intelligence for executive protection: when a model defames your CEO
A model that tells a journalist your CEO is under investigation — when they are not — is a defamation event at scale. DNI is how organisations detect, trace, and correct model-generated falsehoods about their executives.
Deep Narrative Intelligence as competitive intelligence: what models say about your rivals
If you know what AI models say about your competitors — their strengths, their weaknesses, their distortions — you hold a strategic asset. DNI turns the model layer into a source of competitive intelligence.
Risk, by industry.
Finance: why AI assistants must cite verified sources
When a fabricated number can influence an investment decision within seconds.
Healthcare: the risk of clinical hallucination in AI assistants
Why information integrity becomes a patient-safety issue.
Public sector: information sovereignty and generative AI
What institutions must demand from models that answer on their behalf, indirectly.
Energy & industry: protecting reputation against AI narratives
Between a polarised public debate and content-hungry models, how to secure the factual narrative.
Defence: information integrity and language models
Why the sector most exposed to disinformation is also the least ready for LLMs.
AI Information Integrity in the Public Sector | SetRecord
How public institutions can assess AI-generated claims, governance needs, data flows and deployment requirements.
Deep Narrative Intelligence in M&A: the hidden due-diligence gap
Before a merger closes, acquirers audit financials, legal, cybersecurity. Almost none audit what AI models already say about the target. Here is why DNI is becoming a standard step in M&A due diligence.
Deep Narrative Intelligence before an IPO: what models say about you on listing day
On listing day, every journalist, analyst and investor will ask an AI about your company. If the model's answer diverges from your prospectus, you have a problem that no roadshow can fix. DNI closes that gap before the bell rings.
Deep Narrative Intelligence for financial institutions: when a model gets your balance sheet wrong
Banks live and die by trust in their numbers. When an AI model invents, distorts or stale-copies financial figures, it creates a risk that neither PR nor compliance was built to catch. DNI is the layer that does.
Deep Narrative Intelligence for pharma: when AI asserts your drug is unsafe
A model that tells a patient your drug is unsafe — based on a retracted study, a stale label, or a conflation with a different molecule — is a public-health risk and a commercial risk. DNI is how pharma measures and corrects it.
Deep Narrative Intelligence for energy and defence: strategic sectors, strategic narratives
Energy and defence companies operate at the intersection of national security, geopolitical perception, and commercial interest. When AI models distort their narrative, the stakes are strategic. DNI is how these sectors measure and protect their model-layer reputation.
The proof behind the claim.
Provenance: tracing the origin of an AI claim
The central concept of our method — and why a correction without provenance doesn't hold.
ClaimReview and structured credibility signals
How semantic markup helps models tell a verified fact from a rumour.
Building an "AI-ready" fact sheet for your brand
The single source document every model should be able to cite without error.
The disinformation risk score: how it is calculated
Seven variables, one number — the mechanics behind our risk indicator.
Correcting an AI narrative: the 5-step method
From alert to persistence check — the complete protocol, step by step.
Deep Narrative Intelligence: the method that corrects what AI says about you
How SetRecord interrogates AI models with and without web search, quantifies the gap to verifiable facts, and traces the provenance of every erroneous claim — before delivering a sourced correction.
Deep Narrative Intelligence and the training cycle: why correction takes six months, not six days
Models update their deep knowledge every 3–6 months. A correction that works in days on the web layer can take months on the training layer. Here is the full timeline — and why patience is a methodological requirement, not a virtue.
With and without web search: why DNI interrogates in both configurations
A model answers differently with web search enabled and disabled. One reveals the latent narrative baked into the training data; the other reveals the live retrieval layer. DNI measures both, because they are two different risks.
How the AI Image Score evolves over time: reading the curve
The AI Image Score is not a static number — it is a curve. Understanding its shape tells you whether your narrative is stable, improving, regressing, or volatile, and whether your correction programme is working.
Inside a correction cycle: from detection to verified persistence
A correction is not a single act — it is a cycle with five stages, each of which can fail. This article walks through the full cycle, from detecting a distortion to verifying that the correction has persisted across models and time.
Provenance data retention in Deep Narrative Intelligence: what we keep, for how long, and why
DNI generates provenance data — the evidence trail behind every measurement and correction. What we keep, for how long, and why is a governance question that this article answers with full transparency.
Who answers for the answer.
The EU AI Act: what companies need to anticipate
Transparency obligations, indirect reputational risk — what actually changes.
Who is liable when an AI defames your company?
Model publisher, source-content host, distribution platform: the liability chain, explained.
Crisis management in the generative-AI era
Why your crisis communications plan must now include language models.
GDPR and AI: what models know — and shouldn't say — about you
Personal data, the right to rectification: how GDPR applies to a model's outputs.
Building an internal AI-reliability policy (playbook)
Roles, review cadence, alert thresholds: the skeleton of a governance model that lasts.
Deep Narrative Intelligence and European Information Integrity
How the European context shapes AI information integrity, data protection, deployment choices and multi-model monitoring.
Deep Narrative Intelligence under GDPR: how SetRecord audits AI models without breaking the law
Interrogating AI models about a person or organisation is a data-processing activity. Doing it without a legal basis is a liability. Here is how SetRecord built DNI to be GDPR-compliant by construction.
Deep Narrative Intelligence as an instrument of European digital sovereignty
If the narratives that AI models hold about European institutions, companies and policies are shaped outside Europe, sovereignty is incomplete. DNI is the instrument that makes the narrative layer observable and correctable from within.
How to report Deep Narrative Intelligence to your board
Boards understand risk, metrics, and trends. This article shows how to translate DNI findings into the language of board reporting — with the three metrics that actually drive decisions.
The six-month synthesis: how to present DNI results to your executive committee
Every six months, the DNI programme produces a synthesis for the executive committee. This is not a data dump — it is a strategic narrative that connects the AI narrative layer to business outcomes. Here is how to structure it.
See what AI says about you
A SetRecord audit checks what supported AI models say against your verifiable facts — within five business days.