Editor’s take: The search bar as we knew it is dying. Users no longer want 10 blue links—they want direct answers, synthesized from multiple sources, with citations. Perplexity crossed 10 million monthly active users in 2024; Google’s SGE is rolling out globally; and Arc Search’s “Browse for me” feature went viral. The AI disruption in search is real, and the stakes are enormous for publishers, advertisers, and anyone who depends on organic traffic.
The Shift from Links to Answers
Traditional search engines—Google, Bing, DuckDuckGo—return ranked lists of web pages. Users click, skim, and synthesize. AI search engines flip the model: they read the pages for you and return a single synthesized answer with inline citations. According to Gartner, by 2026, 30% of outbound enterprise marketing content will be synthetically generated—and a growing share of search queries will be answered by AI rather than direct traffic to publisher sites.
The behavioral shift is measurable. Perplexity reports that 40% of its queries are research-style questions (multi-hop, comparative, analytical) rather than simple lookups. Users who try AI search tend to stick: retention rates for Perplexity Pro subscribers exceed 70% at 12 months. The question isn’t whether AI will change search—it’s how fast and how far.
Perplexity: The AI-Native Pioneer
Perplexity launched in 2022 as a search engine that answers questions with cited, synthesized responses. By 2026, it has become the default for many researchers, students, and professionals. Key differentiators:
- Real-time web grounding: Perplexity fetches live web results and cites sources inline. Unlike static LLMs, it stays current.
- Pro plan with GPT-4o and Claude: Subscribers get access to frontier models for complex reasoning tasks.
- Vertical expansion: Perplexity has added academic search, code search, and enterprise deployments.
Perplexity’s valuation crossed $3 billion in 2025. The company has partnerships with SoundHound for voice search and is expanding into B2B with Perplexity for Teams. For AI startups building on top of search, Perplexity’s API is a key distribution channel.
Google SGE: Search Generative Experience
Google’s response to AI search is SGE (Search Generative Experience), integrated into the main search results page. When users ask complex questions, an AI-generated snapshot appears above traditional results—pulling from Google’s index and Knowledge Graph. SGE is rolling out globally; by mid-2026, it’s available in 100+ countries and 50+ languages.
The tension for Google is existential: AI answers reduce clicks to publisher sites, which threatens the ad ecosystem. Google is testing ad placements within SGE responses and experimenting with “conversational” search that keeps users in a loop. Early data suggests SGE increases engagement (time on search) but decreases outbound clicks by 15–25% for informational queries—a trend publishers are watching closely.
SearchGPT and OpenAI’s Search Ambitions
OpenAI launched SearchGPT (previously “Search” in ChatGPT) in 2025, enabling ChatGPT to search the web and cite sources. For ChatGPT Plus and Team users, it’s the default for factual queries. SearchGPT uses Bing’s index under the hood (Microsoft’s partnership with OpenAI) and presents answers in ChatGPT’s conversational format.
OpenAI’s strategy is to make ChatGPT the universal interface—search, coding, writing, analysis—all in one place. SearchGPT is a wedge: once users search inside ChatGPT, they’re less likely to leave for Google. Comscore data from early 2026 suggests ChatGPT’s share of “search-like” queries has reached 3–5% in the US, with higher penetration among younger users.
Arc Search: “Browse for Me” and the Mobile Shift
Arc Search, from The Browser Company, took a different approach. Instead of a standalone search engine, it built “Browse for me”—a feature that, when you ask a question, opens multiple tabs, reads the pages, and generates a custom summary. It went viral on social media in early 2025 and has been downloaded millions of times.
Arc’s insight: on mobile, users don’t want to tap through 10 links. They want one answer. “Browse for me” is essentially AI search embedded in the browser. The Browser Company has since added similar features for shopping and research. Arc Search highlights how AI disruption can come from unexpected places—not just search incumbents or AI labs.
Market Size and Revenue Models
The AI search market is still nascent. Perplexity generates revenue from subscriptions (Pro at $20/month) and API access. Google and Bing monetize via ads within AI answers—a format that’s evolving. Arc Search is free; monetization may come from partnerships or premium features.
Estimates vary, but the addressable market for “answer engine” queries could reach $20–30 billion by 2030 if AI search captures 15–20% of query volume. The bigger question is what happens to SEO: if AI answers replace clicks, publishers will need new strategies—perhaps licensing content to AI engines or optimizing for “AI citation” rather than rank.
The Publisher Dilemma
Publishers face a paradox. AI search engines need their content to train and ground answers—but if answers are synthesized in the search result, users never click through. Some publishers have blocked AI crawlers; others are negotiating licensing deals. The outcome will shape the economics of journalism and content creation. Regulatory pressure—e.g., EU rules on AI training data—may force compensation. For now, the relationship is tense and evolving.
Implications for the Broader AI Ecosystem
AI search is a gateway to AI tools adoption. Users who try Perplexity or SearchGPT often expand into other AI use cases—writing, coding, analysis. The open vs. closed source AI debate matters here: open models (Llama, Mistral) power some search backends, while closed models (GPT-4, Claude) power others. Latency and cost drive the choice; for real-time search, smaller, faster models often win.
For more on how AI is reshaping industries, see our coverage of AI in supply chain and AI in legal.
Key Takeaways
- AI search engines (Perplexity, SearchGPT, Google SGE, Arc) are shifting users from links to synthesized answers.
- Perplexity has 10M+ MAU and a $3B+ valuation; retention for Pro exceeds 70%.
- Google SGE reduces outbound clicks by 15–25% for informational queries—publishers must adapt.
- Arc Search’s “Browse for me” went viral, showing AI search can emerge from browsers, not just search engines.
- The AI search market could reach $20–30B by 2030; SEO and publisher monetization are in flux.
For users: The best AI search engine depends on your use case. Research and analysis? Perplexity or SearchGPT. Quick factual lookup? Google SGE or traditional search. Mobile-first, minimal friction? Arc Search. Try several; habits are still forming, and the landscape will evolve. For building AI products that integrate search, consider which engines offer APIs and what licensing terms apply.
Further Reading
Related: VC Fund Structure: GP, LP, Fund Size and Portfolio — The VC Wire
Related: Down Rounds: Impact on Founders, Employees and Investors — The VC Wire
Dive deeper: This article is part of our comprehensive guide — The State of AI in 2026: Everything You Need to Know.