Recent announcements from NucleusLinks, OtterlyAI, Xnurta, InMobi Advertising, Rival Technologies, Luma, and Kineto asked marketing organizations to fund two opposing readers at once. Four of the seven sell instrumentation for the machine that crawls, buys, or transacts against a brand. Two price the human reaction to machine-made output. Rival Technologies, a conversational research technology company, reported on August 13, 2026 that 72% of Gen Z consumers have taken direct action against a brand after encountering AI-generated marketing, based on a July 2026 study of 901 Gen Z participants across the United States and Canada. A team funding both sides of that split works from one budget and two directions for it.
OtterlyAI, an AI search optimization platform, stated the machine-side measurement gap in operational terms on August 13, 2026 when it released Agent Analytics. OtterlyAI’s position is that conventional web analytics runs on JavaScript while AI agents request pages directly from the server, so agent visits never reach the analytics dashboard a marketing team already pays for. Agent Analytics reads server logs instead, reporting visits from named agents including ChatGPT-User, Claude-Bot, Perplexity-User, and Google-Agent, listing the most crawled URLs, and sorting each agent into one of three categories: an on-demand fetcher answering a live user question, a search index crawler, or a training scraper. That third distinction carries the operating consequence. A team that reads a rising crawl count as demand signal, when most of the volume comes from training scrapers, will fund generative engine optimization work against a number that does not move revenue. OtterlyAI stated that it rolled the feature out quietly the week before the announcement and that more than 50 websites connected to it before any public statement, and the company described itself as used by more than 40,000 marketing professionals as of August 2026.
Server-log visibility answers a narrow question and leaves a wider one open. Agent Analytics reports which pages agents fetched. It does not report which prompts produced a citation, and OtterlyAI placed crawl-to-citation conversion rates on a roadmap for later in the quarter rather than in the shipping release. A marketing leader who buys the product this quarter gets an inventory of machine attention and still cannot price it. That gap matters more as zero-click search moves more brand discovery inside answers a team never sees.
Xnurta, an agentic AI advertising platform, took the same premise into retail media on August 13, 2026 and attached governance to it. Xnurta announced general availability of Xnurta MCP, a Model Context Protocol server that exposes retail media data to ChatGPT, Claude, and other enterprise AI platforms, with read-only Data Query shipping first and write capabilities entering closed beta the same day. Xnurta published dates for the phases that follow: campaign optimization enters closed beta on August 11, 2026 and reaches general availability on August 25, 2026, and bulk operations covering bids, budgets, campaign states, keywords, and targets enter closed beta on September 8, 2026 and reach general availability on September 22, 2026. The design choice worth reading is what Xnurta calls action spaces, configurable boundaries that set what an assistant can change. CEO Kashif Zafar framed permission as persistent policy rather than per-action approval. A retail media team adopting that model writes its risk tolerance once, as configuration, and then lives with what it wrote. Someone on that team has to own the boundary file the way an ad operations lead owns a bid cap.
InMobi Advertising, part of InMobi, put a conversational agent on the buy side on August 13, 2026, releasing an agent inside Buyer Hub, its self-serve deal platform, and opening Buyer Hub to agencies and managed service demand-side platforms with role-based access controls, inventory discovery, and self-serve reporting. Chief Business Officer Kunal Nagpal stated that the agent recommends curation and optimization moves while the trader makes the call. InMobi launched Buyer Hub in June 2025 with curators live first, and the company stated that a scoping exercise with agency and managed service DSP customers shaped the version now available. Read alongside Xnurta, the pattern holds across both sides of the transaction: the vendor ships the recommendation and hands the approval back. Agencies evaluating either product should price the review labor, because a system that proposes twenty optimization moves a day creates twenty decisions a human still signs.
NucleusLinks supplied the input layer those systems read. On August 14, 2026, NucleusLinks, the affiliate marketing software platform inside the Affinity Global ecosystem, launched Product Feeds, a data infrastructure layer that standardizes and enriches merchant product data for affiliate publishers. NucleusLinks CEO and Affinity Global CIO Cornelius Frey stated that erroneous product data and pricing that runs even a few hours stale rank among the most damaging conversion killers because they erode reader trust after the click. NucleusLinks named large language model providers first in the list of businesses the launch targets, ahead of loyalty programs, review sites, deal and comparison platforms, and media houses with commerce editorial teams. Publishers running commerce content now maintain product data as infrastructure that answer engines read, which puts feed accuracy on the same footing as page copy. NucleusLinks stated that the platform connects publishers to more than 70 affiliate networks and more than 10,000 merchants across the United States, United Kingdom, European Union, and Asia-Pacific, and that Product Feeds ships with a one-month free trial and no setup fees. Teams building against agentic commerce protocols face the same requirement from a different direction.
Rival Technologies measured what happens when the output of all this machinery reaches a person who recognizes it. In the July 2026 study, Rival reported that 74% of Gen Z respondents react negatively when they realize a brand’s marketing was made with AI and 8% react positively, and that half describe the negative reaction as very negative. The behavioral figures carry more weight than the sentiment ones: Rival reported that 50% of respondents have unfollowed a brand on social media, 49% have complained to friends, family, or online, 48% have unsubscribed from emails or texts, and 43% have stopped buying from a brand altogether. Rival ran the study on its own conversational research platform among participants drawn from its proprietary panels, and Rival sells that platform, so a buyer weighing the numbers carries the vendor authorship into the comparison. Rival also reported that strong negative reactions rose with age inside the Gen Z sample, from 44% among respondents aged 18 to 20 to 54% among those aged 25 to 29, and that 84% of Canadian respondents reacted negatively against a lower United States figure, which tells a North American team that one playbook will underperform in one of the two markets.
The Rival findings do not argue against production tooling. They price disclosure and craft. A Chief Marketing Officer (CMO) who has moved brand social and lifecycle email onto generative production now has a number to model against: 43% of a surveyed Gen Z cohort stopped buying after recognizing AI-generated marketing. Work that scenario against a real book of business. A direct-to-consumer brand with $40 million in annual revenue and 30% of it from customers aged 18 to 29 puts $12 million in the exposed segment. A recognition event that moves even three points of that segment to lapse costs $360,000 in annualized revenue, against a content production line that saves perhaps $200,000 a year in agency and freelance fees. The arithmetic does not forbid AI production. It says the savings sit downstream of the recognition risk, and that the team should spend part of the savings on the editorial pass that keeps the work from reading as machine output.
Luma, a multimodal general intelligence platform, and Dumbstruck, an emotion analytics company, sold that pass as a product on August 13, 2026, announcing a partnership they call Creative Intelligence. Dumbstruck applies AI-powered facial coding to measure how consumers respond emotionally and behaviorally to video advertising before a campaign launches, and Luma generates and refines the assets. Sandro Corsaro, Global Head of Creative and Content at Wayfair, stated that combining Dumbstruck’s emotional intelligence platform with Luma AI lets the creative team understand emotional response and refine creative continuously. The premise underneath the partnership is that generation volume outran selection capability, and that a brand now needs a scoring step between production and media. Buyers should ask what the facial coding panel looks like, because a pre-flight emotion score inherits every limitation of the sample it was trained and tested on.
Kineto, an AI company spun out of JetBrains, released Kinetik For Teams on August 13, 2026, an agent that runs influencer and content marketing from inside Slack and WhatsApp, handling creator discovery and vetting, outreach, brief development, content support, and reporting, and pulling GA4, Amplitude, and Meta Ads data on a weekly or on-demand basis. CEO and co-founder Andrew Zakonov, who created JetBrains’ coding agent Junie, cited eMarketer data that United States influencer marketing spend is set to grow nearly 16% in 2026 and stated that adoption stays concentrated among the largest companies because the operational load is too heavy for smaller teams. Kineto priced subscriptions from $50 a month for teams of 1 to 100 people. At that price a marketing team can run the experiment without a procurement cycle, which is the point and also the risk: creator vetting and outreach are the two places in an influencer program where a bad automated decision reaches a third party under the brand’s name before anyone internal reads it.
The two halves of the week point at the same operating question from opposite ends. Agent-facing infrastructure raises the value of content a machine can parse. Consumer research raises the cost of content a person can identify as machine-made. Teams that treat those as one budget line will underfund one of them.
Strategic priorities for marketing leaders
- Separate agent crawl volume from agent demand signal. Ask any AI visibility vendor to break traffic into on-demand fetchers, index crawlers, and training scrapers before you report the number upward, because the three categories carry different revenue meaning.
- Assign an owner to agent permission configuration. MCP action spaces and equivalent boundary settings are standing policy, and a configuration nobody owns becomes a control nobody reviews.
- Price the human review labor in agentic advertising tools. Recommendation systems that keep the trader in the approval seat move work rather than remove it, and the staffing plan should say who approves and how fast.
- Model recognition risk against your own age mix. Rival’s July 2026 finding that 43% of surveyed Gen Z stopped buying after recognizing AI-generated marketing has a dollar value specific to your revenue concentration.
- Treat product feed accuracy as answer-engine copy. Publishers and brands whose product data feeds AI shopping surfaces now carry stale-price risk into the answers themselves.
- Split United States and Canadian creative testing where Gen Z matters. Rival reported an 84% negative reaction among Canadian respondents, above the United States figure, which breaks single-playbook North American planning.
Press Release Roundup: August 13–14, 2026
NucleusLinks Launches Product Feeds to Close Affiliate Commerce Data Gap NucleusLinks, the affiliate marketing software platform built inside the Affinity Global ecosystem, announced Product Feeds on August 14, 2026, a product data infrastructure layer that helps affiliate publishers standardize, enrich, and activate merchant data at scale. NucleusLinks stated that product information arrives from dozens of merchants, affiliate networks, and feed formats with different structures, update schedules, and quality levels, producing stale prices, missing images, broken stock status, and duplicate listings. NucleusLinks CEO and Affinity Global CIO Cornelius Frey stated that erroneous product data and pricing outdated by even a few hours rank among the most harmful conversion killers because they erode user trust after the click-out. The company named large language model providers, loyalty programs, review sites, deal and coupon platforms, price comparison platforms, and media houses with commerce editorial teams as target buyers, and stated that standardized product data can also serve as a product-level signal for advertiser campaigns across Affinity’s commerce environments. NucleusLinks stated that it connects publishers to more than 70 affiliate networks and more than 10,000 merchants across the United States, United Kingdom, European Union, and Asia-Pacific. Product Feeds is available with a one-month free trial, no setup fees, and no lock-in commitments. Read the full press release (Published: August 14, 2026 | Source: Business Wire)
OtterlyAI Launches Agent Analytics, Making AI Agent Traffic Visible and AI Search ROI Measurable OtterlyAI, an AI search optimization platform, announced Agent Analytics on August 13, 2026, a feature that reports every AI agent and crawler visiting a website, which pages they reach, and how that activity connects to visibility in AI-generated answers. OtterlyAI stated that common web analytics tools run on JavaScript while AI agents request pages directly from the server, so agent visits never appear in analytics dashboards, and that Agent Analytics closes the gap by reading server log data. Inside an OtterlyAI Brand Report, teams see total agent visits, top AI search engines, trends over time, most crawled URLs, a searchable list of every page agents visited, and a catalog of every identified bot, including ChatGPT-User, Claude-Bot, Perplexity-User, and Google-Agent. OtterlyAI categorizes each agent as an on-demand AI fetcher answering a live user question, a search index crawler, or an AI training scraper, so teams can interpret crawl volumes correctly. OtterlyAI stated that it rolled Agent Analytics out to first customers the prior week and that more than 50 websites were already tracking agent behavior before the public announcement. The company described crawl-to-citation conversion rates and deeper funnel tracking connecting agent activity and AI referral traffic to signups and revenue as roadmap items for later in the quarter, and stated that more than 40,000 marketing professionals used its AI search analytics as of August 2026. Read the full press release (Published: August 13, 2026 | Source: GlobeNewswire)
Xnurta Launches MCP for Retail Media, With AI-Powered Campaign Management Entering Beta the Same Day Xnurta, an agentic AI-powered advertising platform, announced general availability of Xnurta MCP on August 13, 2026, a Model Context Protocol server connecting Xnurta retail media data with AI assistants and enterprise AI platforms including ChatGPT and Claude. Xnurta described the launch as the first phase of a roadmap moving AI assistants from answering retail media questions to taking governed actions, with read-only Data Query generally available and the first write capabilities entering closed beta the same day. Xnurta stated that most advertising APIs organize around campaigns, ad groups, targets, and keywords, and that its platform adds objective context by organizing campaigns into managed groups, product lines, and shared objectives that Xnurta MCP can read, so an advertiser can ask how a launch performs against its target rather than identifying every campaign tied to it. As MCP expands from analysis to execution, Xnurta is introducing action spaces, configurable boundaries determining what an AI assistant can change, establishing permissions as persistent policy rather than per-action approval. Xnurta published a phased schedule: campaign optimization enters closed beta on August 11, 2026 and reaches general availability on August 25, 2026, and campaign bulk operations covering bids, budgets, campaign states, negative keywords, keywords, and targets enter closed beta on September 8, 2026 and reach general availability on September 22, 2026. CEO Kashif Zafar stated that marketers are already moving work into general-purpose AI assistants and should not have to move it back into another interface to manage retail media. Read the full press release (Published: August 13, 2026 | Source: Business Wire)
InMobi Advertising Launches A Conversational AI Agent for Media Buyers InMobi Advertising, part of InMobi, announced a conversational AI agent on August 13, 2026, powered by the company’s unified intelligence layer and running inside Buyer Hub, its self-serve deal platform. The agent helps media buyers discover inventory and supports building deals and optimizing campaigns through natural language. InMobi also opened Buyer Hub to agencies and managed service demand-side platforms, providing direct access to InMobi ad inventory with role-based access controls, inventory discovery that lets teams curate supply before activation without email exchanges, and self-serve reporting that surfaces campaign metrics and diagnostics directly. Chief Business Officer Kunal Nagpal stated that agencies had been asking for direct self-serve access since InMobi opened Buyer Hub to curators the prior year, and that the company built the agent so a trader does not have to choose between speed and control, with the agent recommending curation and optimization moves while the trader makes the call. InMobi launched Buyer Hub in June 2025 with curators live first and agency access named as the next phase, and stated that a market requirement scoping exercise with agency and managed service DSP customers shaped the product now live. The launch follows InMobi’s partnership with Scope3, which opened premium connected TV and mobile in-app supply, including Glance, to AI-powered buying agents on Scope3 Interchange. Agencies and managed service DSPs receive access through sales-assisted onboarding. Read the full press release (Published: August 13, 2026 | Source: Business Wire)
The AI Backlash: Nearly Three in Four Gen Z Have Punished a Brand Over AI Marketing, New Rival Technologies Study Shows Rival Technologies, a research technology company specializing in conversational research, released The Gen Z AI Backlash: How Young Consumers Really Feel About AI in Marketing on August 13, 2026. Rival reported that 72% of Gen Z consumers have taken direct action against a brand after encountering AI-generated marketing, including unfollowing, unsubscribing, complaining publicly, and abandoning a purchase. Rival reported that 74% of respondents react negatively when they realize a brand’s marketing was made with AI, with half describing that reaction as very negative, and that 8% react positively. On behavior, Rival reported that 50% have unfollowed a brand on social media, 49% have complained to friends, family, or online, 48% have unsubscribed from emails or texts, and 43% have stopped buying from a brand altogether. Rival reported that strong negative reactions rose with age within the sample, from 44% among respondents aged 18 to 20 to 54% among those aged 25 to 29, and that 84% of Canadian respondents reacted negatively, above the United States figure. Rival stated that the strongest objections tied less to whether AI-generated content looked convincing and more to what respondents believed AI meant for people, including job displacement and effects on artists and creative workers. Rival conducted the study in July 2026 on its AI-accelerated, mobile-first conversational research platform among 901 Gen Z participants across the United States and Canada, drawn from Rival’s proprietary panels. Read the full press release (Published: August 13, 2026 | Source: PR Newswire)
Luma and Dumbstruck Launch Creative Intelligence, the Missing Decision Layer for Advertising Luma, a multimodal general intelligence platform, and Dumbstruck, an emotion analytics company, announced a partnership on August 13, 2026 to define and build what the companies call Creative Intelligence, applying AI and measured human response to advertising content itself. The companies stated that brands can now generate more creative, more variations, and more versions than before, and that the constraint has moved to knowing which content will earn attention, drive emotion, and perform before it enters the market. Dumbstruck uses AI-powered facial coding to measure how consumers respond emotionally and behaviorally to video advertisements before a campaign launches, and the two technologies form a loop in which Dumbstruck identifies where content can improve and Luma Agents produce the refined assets. Sandro Corsaro, Global Head of Creative and Content at Wayfair, stated that combining Dumbstruck’s emotional intelligence platform with Luma AI lets his creative team better understand emotional response and continuously refine creative while maximizing production and media investment. Dumbstruck CEO Jeff Tetrault stated that AI is changing the economics of advertising creative. Read the full press release (Published: August 13, 2026 | Source: Business Wire)
Kinetik Launches For Teams: an AI Team Agent That Manages Social and Influencer Marketing Campaigns Kineto, an AI company spun out of JetBrains, launched Kinetik For Teams on August 13, 2026, an AI agent that runs influencer and content marketing from inside Slack and WhatsApp. Kineto stated that the agent joins a team’s chat, learns the company once, and handles creator identification and vetting including reach, relevance, brand fit scoring and fake engagement flagging, personalized outreach and negotiation, content writing in brand voice for search and AI outputs, and campaign analytics pulling GA4, Amplitude, and Meta Ads data weekly or on demand. The agent also monitors competitors and trending topics across X, Instagram, TikTok, and news sources. CEO and co-founder Andrew Zakonov, who created JetBrains’ coding agent Junie, stated that United States influencer marketing spend is set to grow nearly 16% in 2026 according to eMarketer, and that adoption remains concentrated among the largest companies because the operational load is too heavy for smaller teams. Andrew Koval, co-founder of Meadow AI, stated that the agent analyzes creator content across platforms, ranks topics by performance, surfaces median metrics, and performs sentiment analysis on product perception. Kineto stated that Kinetik serves 10,000 creators and priced subscription plans for teams of 1 to 100 people starting at $50 a month. Read the full press release (Published: August 13, 2026 | Source: PR Newswire)
Frequently asked questions
What marketing technology and AI news was announced on August 13 and 14, 2026? On August 13, 2026, OtterlyAI released Agent Analytics for AI agent and crawler traffic, Xnurta made Xnurta MCP generally available for retail media, InMobi Advertising launched a conversational AI agent inside Buyer Hub, Rival Technologies published its Gen Z AI Backlash study, Luma and Dumbstruck announced a Creative Intelligence partnership, and Kineto launched Kinetik For Teams. On August 14, 2026, NucleusLinks launched Product Feeds for affiliate publishers. Wire volume on August 14 ran light, so this edition covers both days.
What did Rival Technologies find about Gen Z and AI-generated marketing? Rival Technologies reported on August 13, 2026 that 72% of Gen Z consumers have taken direct action against a brand after encountering AI-generated marketing, that 74% react negatively when they realize marketing was made with AI, and that 8% react positively. On behavior, Rival reported that 50% have unfollowed a brand, 49% have complained to friends, family, or online, 48% have unsubscribed from emails or texts, and 43% have stopped buying from a brand. Rival conducted the study in July 2026 among 901 Gen Z participants across the United States and Canada on its own conversational research platform, drawn from its proprietary panels.
What is OtterlyAI Agent Analytics and what does it measure? OtterlyAI released Agent Analytics on August 13, 2026 to report every AI agent and crawler visiting a website using server log data rather than JavaScript-based analytics, which OtterlyAI states cannot record agent visits because agents request pages directly from the server. The feature reports total agent visits, top AI search engines, trends over time, most crawled URLs, and a catalog of identified bots including ChatGPT-User, Claude-Bot, Perplexity-User, and Google-Agent. OtterlyAI categorizes each agent as an on-demand AI fetcher, a search index crawler, or an AI training scraper. OtterlyAI stated that more than 50 websites connected to the feature during its quiet rollout the prior week.
What did Xnurta announce with its MCP server for retail media? Xnurta announced general availability of Xnurta MCP on August 13, 2026, connecting retail media data with AI assistants including ChatGPT and Claude, with read-only Data Query available and the first write capabilities entering closed beta the same day. Xnurta is introducing action spaces, configurable boundaries determining what an AI assistant can change, which establish permissions as persistent policy rather than per-action approval. Xnurta published a schedule placing campaign optimization at general availability on August 25, 2026 and campaign bulk operations at general availability on September 22, 2026.
What should marketing leaders take from the August 13 and 14, 2026 releases? Four of the seven releases sell instrumentation for machine-mediated demand, covering agent crawl visibility, retail media data access for AI assistants, conversational media buying, and standardized product feeds that answer engines read. Two price the human reaction to machine-made output, with Rival Technologies reporting on August 13, 2026 that 43% of surveyed Gen Z stopped buying from a brand after recognizing AI-generated marketing, and Luma and Dumbstruck selling pre-launch emotion scoring. Marketing leaders funding both agent-facing optimization and human-facing craft should budget them as separate lines, because gains on one side do not offset losses on the other.
