A platform--creator mechanism called VCR based on verifiable-content rewards, which credits rewrites that surface checkable factual substance, aligning creator incentives with answer trustworthiness and produces a win--win outcome under the empirical equivalence criterion.
Abstract
Generative engines are reshaping the web ecosystem by making citations a key mechanism for allocating attention, attribution, and downstream value. This creates a strategic tension: content providers are incentivized to optimize for model citation, while platforms must preserve answer quality and trustworthy attribution. We show that this tension can escalate into citation wars. In repeated simulations, state-of-the-art generative engine optimization (GEO) attacks adapt to conventional defenses by producing citation-seeking rewrites that degrade document quality and introduce unsupported claims. To study this problem, we formulate the supplier--platform interaction as a repeated Stackelberg game with partial monitoring. A local best-response analysis identifies when citation competition approaches an inert stationary outcome. Motivated by this finding, we propose a platform--creator mechanism called VCR based on verifiable-content rewards. Rather than only penalizing suspicious rewrites, the platform also credits rewrites that surface checkable factual substance, aligning creator incentives with answer trustworthiness. Experiments on three benchmarks show that VCR consistently achieves the largest Net defense-utility score, outperforming the strongest baseline by an average of 12.1 percentage points, and produces a win--win outcome under our empirical equivalence criterion.
This paper focuses on defending generative search engines against malicious Generative Engine Optimization (GEO), which rewrites web documents to match engines'citation preferences and thereby manipulates generated answers. Recent GEO methods have advanced from hand-crafted rewriting to automated and agentic optimization, substantially increasing the visibility of target documents in generated answers. However, defending against such manipulation poses two major challenges: attack documents remain factually consistent with their originals, rendering fact verification and perplexity filtering ineffective, and the features they amplify equally characterize high-quality benign content. To address these limitations, we propose GEO Defender, a two-stage defense aligned with the attack chain that requires no fine-tuning of the target LLM. GEO Defender consists of Shield Reranker and Training-Free Shield Generation (TFSG). Specifically, Shield Reranker learns a preference-based defensive residual over a frozen base reranker, demoting GEO-rewritten documents while preserving relevance judgments, and TFSG distills defense outcomes into a natural-language experience library that guides the target LLM's source use at inference. Experiments on two state-of-the-art closed-source LLMs and three open-source LLMs across seven GEO attacks demonstrate that GEO Defender reduces the average attack success rate from 50.32% to 6.20%, retains 94.12% of benign-evidence use, preserves answer quality, and generalizes to unseen attacks from construction instances.
Hao-Zhang Li, Yang-Guang Shao, Xin-Jie Lin et al.· 0 citations
This study provides an empirical and theoretical investigation into the systemic inefficiencies confronting contemporary programmatic advertising architectures: namely, capital inflation driven by over-reliance on short-term attribution (the "Performance Trap") and the conversion friction imposed on non-established market entrants (the "Obscurity Tax"). We examine the structural realignment of information retrieval pipelines as consumer search behaviors migrate from legacy index-based search engine results pages toward conversational, Large Language Model (LLM) interfaces. Utilizing multi-channel Google Analytics 4 (GA4) telemetry, this paper quantifies the operational divergence between active brand equity and anonymous enterprises within identical commercial verticals.
To mitigate these barriers, we present a full-cycle data engineering architecture developed at iLION Digital that leverages server-side Google Tag Manager (GTM), Google BigQuery, and custom SQL identity resolution scripts to operationalize automated Value-Based Bidding (VBB) via direct API synchronization. A cross-vertical meta-analysis of five multi-market enterprise accounts across the E-commerce, Medical, and Home Services industries validates the scalability of this infrastructure, demonstrating an aggregate increase in inbound inquiries of 220.10%, a contraction in average Cost-Per-Acquisition (CPA) to 80.15% of historical baselines, a 225.42% expansion in total transactional value, and a net increase in Return on Investment (ROI) of 124.45%.
Vladyslav Bilinchuk, Oleksandr Korogovnyi· Metaverse Science, Society a...· 0 citations
The rise of generative search engines and large language model assistants has shifted digital visibility from clicking a link to being present within synthesized answers, a change that traditional search engine optimization (SEO) metrics fail to capture. This study presents a scoping review, conducted following JBI methodology and reported according to PRISMA-ScR, aimed at mapping how the emerging literature on Generative Engine Optimization (GEO) measures the presence, visibility, and citation of sources and brands. The search combined an AI-assisted discovery tool, citation chasing, and reproducible multilingual queries in OpenAlex, with verification in an indexed database. Twenty-three primary studies were included (2023-2026), mostly preprints, plus two reviews registered as a conceptual framework. Findings reveal a nascent, fragmented field: seven partially overlapping metric families with no integrating framework; a heterogeneous unit of analysis migrating from the source/URL toward the brand and the agent trajectory; insufficient control of engine stochasticity; and, most critically, the absence of psychometrically validated instruments, as measurement relies on ad hoc benchmarks rather than validated tools. We conclude that no integrated, validated instrument for generative brand presence yet exists, and that this gap —in construct validity and documented reliability— justifies developing one. The study also offers a methodological lesson on the under-coverage of indexed databases in this born-digital field.
Néstor Romero Ramos, Yulianna Lobach, David Abreu Abreu et al.· Latitude· 0 citations
Social media runs on sharing; copyright runs on exclusion. When these operating systems collide, platforms rewrite the rules through click-through contracts, opaque algorithms, and sweeping licenses. This article traces that shift: from paparazzi lawsuits to embedding fights, from US “server test” chaos to the CJEU’s move toward platform-as-publisher. It unpacks the DSM Directive’s bargain –
ex ante
“best efforts” (filters, licensing, stay-down) versus hard-edged safeguards for lawful uses – and shows how over-blocking and private Terms of Service still tilt the field. The payoff is a practical blueprint: machine-readable signals for exceptions, auditable error targets, symmetric notice/counter-notice, non-override of users’ rights, targeted injunctions, fairer collective licensing, and real EU-level harmonization. Bottom line: calibrate code, contracts, and courts so creativity thrives without turning platforms into copyright police.
Luca Ettore Perriello· IIC - International Review o...· 0 citations
This study examines how platform-mediated environments reshape the temporal structure of business communication cascades and alter CEO sensegiving authority during technology disruption events.
Analysing the DeepSeek R1 episode of January to February 2025, we apply process tracing and directed content analysis to 33 purposively sampled communications from 25 named actors across developer, influencer, market, political, and executive communities. Both coders independently coded all 29 codable events; intercoder reliability was assessed via Cohen’s kappa (acknowledgement κ = .76, reframing κ = .85, defensive minimisation κ = .89, geopolitical κ = .91), ranging from substantial to almost perfect agreement.
We identify an eight-phase reverse communication cascade in which developers, elite influencers, and financial markets established interpretive frames before incumbent CEOs responded, inverting the conventional CEO-led sequence assumed in agenda-setting and organisational communication research. The study introduces reputational adjacency to describe a crisis-like condition in which reputational pressure on a firm arises from a competitor’s technological success rather than organisational failure. Findings reveal a consistent acknowledge-reframe dyad as the dominant CEO communication response.
These concepts extend business communication and crisis communication theory by specifying how platform logics structurally constrain executive sensegiving in competitive disruption contexts.
R. K. Gupta, V. Vohra, Neerja Pande· International Journal of Bus...· 0 citations
HERALD is introduced, an offline audit that applies exact same-question interventions, separates candidate-visible from oracle information, and enumerates detector contracts before policy optimization, and separates robust scoring, sparse learning signal, and policy transfer.
Zhuowen Liu, Bohan Cui, YinShang Guo et al.· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
