AIVO Journal

Profound published a 3,000 word comparison of itself against AthenaHQ this week. It is the wrong comparison. Here is the gap both platforms share. There is a turn in every AI buying conversation that the entire GEO and AEO category has collectively decided not to measure. It happens when

A peer-archived study of five major brands reveals a structural gap between AI visibility and AI purchase recommendation outcomes The GEO and AEO category was founded on a correct insight. Traditional search metrics don't measure brand performance in AI-generated responses. Citation frequency, share of AI voice,

When a mass market brand loses the AI recommendation, it is usually a content architecture problem - the model can't find the structured evidence it needs to pass the criteria filter. Fix the evidence layer, improve the position. For luxury brands the problem runs deeper. The brand equity that

The channel is now accessible to the buyers who care most about attribution. The measurement infrastructure they need is not there yet. ChatGPT advertising launched in February 2026 as an enterprise CPM product. Ten weeks later it has a self-serve ads manager, CPMs down from $60 to $25, minimum

How organic inference position became the most important variable in media planning — and what agencies need to measure before a single dollar enters ChatGPT inventory The brief that nobody can answer yet Every agency holding company is fielding the same client question in 2026: should we be on ChatGPT advertising,

AIVO Evidentia · Working Paper WP-2026-04 · April 2026 DOI: 10.5281/zenodo.19588522 Abstract This article announces the publication of brand.context — a machine-readable brand context standard designed for consumption by AI agents during commerce and purchase recommendation tasks. brand.context is the implementation layer of the AIVO

Why the brand you can see in AI is not the brand that wins the sale Every week, brands across the world pay for tools that tell them how often AI mentions them. Share of voice. Citation frequency. Mention rates. The dashboard updates. The numbers go up. The quarterly report

A CODA score of 8/100 on the day of acquisition. Here is what that means. On the day Unilever announced the acquisition of Grüns for $1.2 billion, we had fresh probe data from a buying sequence we had run that morning. We did not plan the timing. We

Institutional Representation Risk in Financial Services Abstract As large language models increasingly mediate product and institutional selection in retail financial services, representation outcomes have become non-uniform across AI systems and unstable over time. Identical queries can produce divergent institutional winners across models, while periodic model updates alter survival to

Abstract The Q1 2026 Global Banking AI Decision Index demonstrated that large language model recommendation systems exhibit measurable structural concentration, late-stage substitution, and cross-platform instability. Subsequent industry response and media coverage elevated AI-mediated selection into executive discourse. This article argues that the decision layer now requires formal

Evidence from the Global Banking AI Decision Index (Q1 2026) and Subsequent Industry Response Authors: Tim de Rosen, Paul Sheals Affiliation: AIVO Evidentia Ltd Date: February 2026 Abstract The AIVO Surface™ Global Banking AI Decision Index (Q1 2026) evaluates how major banking institutions perform at the decision stage within large

The Infrastructure Layer for AI-Mediated Purchase Decisions 58% of buyers now use AI systems to choose between competing brands. Not to research. Not to browse. To choose. “Which CRM should we buy?” “Book me a hotel in Paris.” “What’s the best anti-aging serum?” AI systems are answering

Digital leaders have spent two decades optimizing for visibility. Rank position. Click-through rate. Conversion rate. Share of search. But AI assistants are introducing a structural shift. They do not distribute attention across ranked links. They compress category decisions into a single recommendation. That compression changes competitive dynamics. Presence Is

If an AI assistant generates a materially adverse or misleading representation of your institution today… Could you reconstruct it six months from now? Precisely. • Exact prompt • Exact model and version • Exact output • Exact timestamp • Demonstrable testing conditions Or would you rely on screenshots and internal recollection? In controlled testing, we

When the U.S. Food and Drug Administration formalized its risk-based credibility framework for AI in healthcare, commentary focused on compliance cost and innovation friction. That interpretation understates the shift. The FDA normalized AI as regulated decision infrastructure. Once AI is treated as infrastructure, accountability expands beyond what an

A new variable is entering forecast models. It is not yet measured. In most institutions, revenue forecasting rests on a stable identity: Expected Revenue = Pipeline Volume × Win Probability × Average Deal Value Pipeline is tracked. Deal value is modeled. Win probability is stress-tested. What is rarely examined is whether the

1. The Evidence Is No Longer Theoretical Under controlled, repeatable prompt classes across major AI systems: * Institutional ordering diverges * Narrative framing shifts under identical queries * Final recommendation resolution varies by model * Displacement patterns remain stable within execution windows This is not anecdotal. It is observable and reproducible. The phenomenon exists.

Abstract Where Case Study 1 documented platform-dependent divergence in oncology, this case documents cross-platform convergence in cardiology. Under identical decision-stage prompts, multiple AI systems consistently positioned the same therapy as preferred. The relevance lies not in clinical deviation, but in authority formation and the visibility of its

Platform Divergence and Decision-Stage Instability in Tier 1 Banking AIVO Journal — Banking Case Series Abstract Controlled cross-model testing across Tier 1 banking institutions demonstrates that identical comparative prompts can yield materially different decision-stage recommendations across leading AI systems. Under fixed prompt sequences and repeated runs, we observe:

Abstract Conversational AI systems are increasingly used at the point of therapeutic choice in oncology. In structured decision-stage testing across multiple leading AI systems, identical treatment-selection questions produced materially different “preferred therapy” outcomes depending solely on platform. The divergence did not arise from guideline deviation or factual error.

The Governance Blind Spot in External AI Systems AIVO Journal – Governance Commentary Abstract AI governance has concentrated on execution risk: whether agents act lawfully, whether decisions are compliant, whether liability can be assigned. This focus leaves a structural blind spot. Before execution, there is representation. Large language models shape how

AI assistants are compressing vendor evaluation into decision workflows that many organisations do not monitor. Over the past two years, SaaS competition has quietly shifted layers. The dominant question used to be: Are we visible? The emerging question is: Are we eligible? Structured multi-model observation across procurement-like prompt

Context The previous analyses in this series established two related problems: external AI systems now generate decision-relevant representations without organisational visibility, and optimisation-first approaches contaminate the evidentiary record organisations later need to evaluate their own interventions. The governance implication follows directly: Observation must precede intervention. The next question

Most optimisation strategies begin from a simple premise: if AI systems are shaping decisions, then influencing those systems is the rational response. What this framing overlooks is that intervention changes the very environment organisations later attempt to evaluate. Once optimization begins: * the original answer state is lost * subsequent outputs are

Abstract Recent analyses of large-scale citation behavior in conversational AI systems have suggested that visibility within citation graphs offers a proxy for influence and decision relevance. This article explains why that assumption fails in decision-adjacent contexts, and why reliance on citation data creates a false sense of governance