We're testing new ad formats in Search and expanding our Direct Offers pilot

Original Article ↗ Hacker News Discussion ↗

Google’s move to embed new ad formats into its AI-powered Search “AI Mode” is widely seen as an inevitable attempt to protect its core ads business as LLM-style answers reduce traditional clicks. Commenters worry that adverts blended into conversational answers will be harder to detect, risk biasing or corrupting information (including for politics and health), and may skirt rules requiring clear labeling of sponsored content. Many say this will further degrade search quality, drive more users toward ad blockers and paid alternatives like Kagi or DuckDuckGo, and deepen the broader trend of the web becoming pay-to-play and dominated by opaque recommendation algorithms.

Business motives and inevitability of AI ads

  • Many see this as inevitable: Google is fundamentally an ad company and search ads are its primary revenue.
  • Commenters argue AI overviews threaten the click-based ad model, so inserting ads into AI answers is “protect the golden goose,” not a surprise.
  • Some think Google is moving too fast and could have waited for OpenAI to “blink first” to avoid reputational damage; others say first-mover advantage with advertisers matters more.
  • Debate on whether Google’s stock is driven by fundamentals vs “vibes,” but general agreement that ad monetization pressure is intense.

Impact on search quality, trust, and bias

  • Strong concern that AI answers with embedded ads will be less transparent than traditional search: you can inspect ranked results, but not a synthesized answer.
  • Fears that advertiser influence will bias AI responses (“best tool for X” becomes pay‑to‑play), especially for products and politics.
  • Some note SEO and affiliate marketing already heavily bias search and LLM training data, so “neutral truth” was never there to begin with.
  • Others say this crosses a line: if the model context or training is quietly influenced by ad spend, users can’t distinguish paid from organic.

Legality and disclosure

  • Multiple references to FTC/consumer rules requiring “clear and conspicuous” ad disclosure; skepticism that enforcement will keep up or be meaningful.
  • Speculation that platforms will use vague global disclaimers like “may include sponsored content,” which may technically comply while staying opaque.

User behavior, ad blocking, and alternatives

  • Many plan to block AI overviews or stop using Google entirely, switching to Kagi, DuckDuckGo, or other engines; some already have.
  • Tips shared: using udm=14 to get “Web” results, browser extensions/userscripts to hide AI overviews, and custom ad/AI blocklists.
  • Some argue users rarely switch platforms for ads; others counter that 30% adblock use is historically huge and suggests rising intolerance.

Are ads ever “helpful”?

  • Most frame “helpful ads” as marketing spin; ads primarily help advertisers and platforms.
  • A minority share cases where targeted ads led them to genuinely useful products or events, especially on Instagram or niche contexts, but call these exceptions.

Broader worries: manipulation and “enshittification”

  • Strong anxiety about AI as a “perfect propaganda” or sales tool, especially for political ads and subtle behavioral nudging.
  • Many see this as another stage of “enshittification” of both search and AI tools, accelerating the decay of the open web.