Google is starting to show what the paid-search playbook looks like when the search results page becomes a conversation. In a May 2026 announcement from Google, the company confirmed it’s testing new Gemini-built ad formats inside AI Mode — including Conversational Discovery ads and Highlighted Answers — along with an “independent AI explainer” displayed inside the ad unit itself (Google Blog).

Why this matters right now (July 2026): AI Mode is not just changing organic visibility. It’s changing how people evaluate options, how ads get selected, and what “good” creative looks like when Gemini is generating contextual copy on the fly.

What Google is testing: Conversational Discovery Ads, Highlighted Answers, and AI explainers

Google says it’s experimenting with two new ad types built with Gemini in AI Mode: Conversational Discovery ads and Highlighted Answers (Google Blog). These are designed to feel like they belong inside an AI-generated response, not bolted onto the page.

  • Conversational Discovery ads aim to answer a user’s specific, nuanced question directly, with creative generated to match the context of that “conversation” (Google Blog).
  • Highlighted Answers make it possible for ads to appear inside AI Mode recommendation lists (for example: “best language apps”) when they’re deemed highly relevant and high quality (Google Blog).
  • Both formats are expected to include an independent AI explainer that “evaluates and synthesizes” product/service information alongside advertiser messaging, to improve transparency and trust (Google Blog).

Google also describes two upcoming “next-gen” Search ad experiences: AI-powered Shopping ads (Gemini writes a custom explainer for a product based on the shopper’s intent) and Business Agent for Leads (a chat-based lead experience inside an ad, trained on the advertiser’s website) (Google Blog).

The hidden shift: your creative becomes a data source for Gemini (not a finished asset)

In classic PPC, you ship a final headline/description set and then optimize. In AI Mode, Google is signaling a different model: Gemini interprets the query context and constructs creative that fits that moment (Google Blog).

That changes what top-performing advertisers should invest in:

  • Structured product/service detail (feeds, schema, spec pages, comparison tables) so Gemini can pull accurate facts.
  • Proof assets (reviews, certifications, warranty terms, shipping/returns clarity) that can withstand an “independent AI explainer.”
  • Clear differentiation (who it’s for, who it’s not for) so the model has guardrails to generate relevant messaging.

If this continues rolling out, the “best ad copywriter” might increasingly be the one who designs the best information architecture behind the copy.

Measurement and governance: why AI Mode ads raise the bar for compliance

Google frames these formats as more transparent: the AI explainer provides context “alongside the advertiser’s creative” and the units remain clearly labeled as Sponsored (Google Blog). But from an operator’s perspective, there’s a new governance problem:

  • If Gemini assembles context-specific ad creative, you need monitoring to ensure claims remain compliant (especially in regulated categories).
  • You’ll want to align internal stakeholders on approved claims and disallowed claims — and ensure landing pages, feeds, and help docs match.
  • Brand safety shifts from “what did we write?” to “what can the system infer and generate from our inputs?”

Practical move: treat your website and product data as if they were part of the ad creative review process — because in AI Mode, they are.

What to do next: a 30-day action plan for agencies and in-house teams

These tests will evolve, but the direction is clear: Search is becoming AI-native, and paid placements are becoming conversation-native. Here’s a concrete playbook to execute now:

  • Audit your “truth layer.” Identify the pages/feeds Gemini would rely on (pricing, specs, policies, reviews). Fix contradictions.
  • Upgrade landing pages for AI comprehension. Add scannable sections, FAQs, comparison blocks, and schema markup so machines (and humans) can verify claims quickly.
  • Strengthen your offer strategy. Google is expanding Direct Offers with promotion bundling, native checkout for UCP merchants, and travel deal integrations (Google Blog). Your promotions need guardrails and profitability logic, not just coupon codes.
  • Reframe creative testing. Test inputs (product descriptors, benefits hierarchy, proof points) that an AI can recombine — not just static headlines.
  • Prepare reporting for “assistant-led” journeys. Expect higher variance in query paths; focus on lead quality, assisted conversions, and downstream revenue.

Bottom line

Google is testing a future where ads don’t just match keywords — they participate in the reasoning layer of Search. If you want to win in AI Mode, treat your marketing as a system: clean data, credible proof, clear differentiation, and measurement that can handle non-linear journeys.

If you want help adapting your SEO, PPC, and content strategy to AI Mode and generative search, Real Internet Sales can help. Call 803-708-5514 or visit realinternetsales.com.