Last updated September 2026
The Authenticity Backfire: How Americans Feel About Marketing
B2B Marketing Strategy • August 30, 2026
A SponsorCX survey of 1,000 U.S. adults found 78% feel modern marketing is more performative than genuine, and 67% say brands overuse the word “authentic” without backing it up. The research, released as an infographic, also covers the enduring pull of word-of-mouth and which brand partnerships actually build audience trust. For marketers leaning on authenticity messaging or creator partnerships, the data suggests overused authenticity claims may now backfire, making concrete proof points more valuable than the word itself.
Google Making AI Overviews Into AI Mode Responses
SEO & AI Search • August 28, 2026
Google confirmed it is testing a version of AI Overviews that loads the full AI response and an open “Ask anything” box by default, without requiring a “Show more” click, which pushes traditional organic results further down the page for affected queries. Google says the dynamic expansion only triggers where its systems judge it “most useful” and cancels if a user has already begun scrolling past the AI Overview. The behavior was first spotted publicly on August 27 and confirmed by Google the next day. This is a further step toward making AI Mode-style answers the default search experience, adding pressure on organic visibility and click-through rates for affected queries.
AI Update, August 28, 2026: AI News and Views From the Past Week
AI in Marketing (Roundup) • August 28, 2026
MarketingProfs’ weekly AI roundup compiles over 20 developments from August 21-28, including Brainlabs data showing AI-referral traffic converts at roughly 1.5x the organic-search rate despite a 10.5% organic traffic decline across 54 advertisers, a study linking OpenAI publisher licensing deals to a 48% jump in ChatGPT citation rates, and Google’s John Mueller stating that solid SEO fundamentals are sufficient for AI search visibility without separate GEO tactics. It also covers The Trade Desk’s new agentic Kokai Zuma platform, X’s launch of an Ads MCP server for third-party AI tools, and IAB’s early work on an AI-attribution measurement framework. This is a broad recap rather than one discrete story, included here as an index to several developments worth monitoring.
ChatGPT Ads adds automated bidding, platform targeting
Paid Media • August 24, 2026
ChatGPT Ads introduced a “Maximize results” automated bid strategy that lets OpenAI’s system set and adjust bids toward a campaign’s chosen goal, shifting bid control from the advertiser to the platform. The update also adds platform-level targeting (advertisers can now restrict delivery to iOS app, Android app, or web), one-day view-through conversion reporting at the campaign, ad group, and ad level, and a new WorkMagic integration for cross-channel reporting and Conversions API signal-sharing. The rollout coincides with ChatGPT Ads expanding into Brazil and Mexico. Advertisers should watch how Maximize results performs against manual bidding and separate view-through conversions from click-driven ones when evaluating campaign performance, since the newly visible attribution can inflate reported results.
Google unifies Google tag and Tag Manager, adds no-code event tracking
Analytics & Measurement • August 24, 2026
Google is merging Google tag and Google Tag Manager into one measurement system, upgrading existing Google tags into full Tag Manager containers with visual, no-code event tracking. Advertisers can now define conversions and events by clicking through their site rather than writing code, while optimized containers send data directly to destinations like Google Ads and Analytics instead of loading extra gtag.js, which Google says cuts measurement latency. The redesign also adds a visual data-flow map and new partner integrations with Akamai, Cloudflare, Duda, Fastly, and Webflow. Advertisers with complex existing Tag Manager setups should preview Google’s proposed optimization changes before publishing, since a simplified configuration may not behave identically to a customized one.
AI-Generated Images Can Perform as Well as Stock Photography
AI & Design / CRO • August 21, 2026
Nielsen Norman Group tested 77 U.S. participants on a fictional consulting firm’s webpage using either AI-generated or real stock hero images, without disclosing which was which, and found no statistically significant trust, professionalism, or authenticity penalty for AI imagery — authenticity actually scored 0.4 points higher on a 7-point scale. Diversity and representation in the image mattered more to participants than whether it was AI-generated or real. Participants who suspected an image was AI-generated did rate the page less favorably, which matters as EU AI Act disclosure rules and similar transparency requirements expand. NN/G recommends evaluating each AI image individually for representation, realism, and visual errors rather than assuming AI images are inherently worse.
How to Avoid Claude Watermarking Your Content
AI & Compliance • August 2026
Anthropic has begun embedding invisible statistical watermarks in Claude-generated text and signed C2PA provenance metadata into supported file outputs, in response to the EU AI Act’s Article 50 transparency requirement that took effect August 2, 2026. The marking applies across Claude.ai, Claude Code, Cowork, Tag, the API, and all cloud platforms worldwide, not just to EU users, and the text watermark persists through copy-paste and some editing. Anthropic has not released a public detector, so marketers publishing Claude-assisted content have no reliable way to check what a third-party AI-detection tool might flag; heavy editing, paraphrasing, or translation can make the text watermark undetectable, while file-level metadata can be stripped by re-exporting or re-saving. The bigger operational risk isn’t the watermark itself but automated plagiarism or AI-detection tools misreading a “detected” mark as proof of full AI authorship, when it can just as easily reflect a light AI proofread of human-written work.