The news: AI platforms are no longer a side note in discovery—they’re driving measurable web traffic. Previsible’s 2025 AI Traffic Report shows that sessions driven by large language models (LLMs) surged 527% in just five months, per Search Engine Land.
AI’s rise is reshaping how users find brands and information as web traffic declines, demanding an immediate strategic response from marketers.
Discovery is shifting to AI: Total AI-referred sessions from OpenAI’s ChatGPT, Perplexity, Anthropic’s Claude, Google’s Gemini, and Microsoft’s Copilot exploded in 2025, jumping from 17,076 in January to 107,100 by May.
- Legal, finance, health, insurance, and small to medium-sized businesses (SMBs) made up 55% of all AI-driven traffic, demonstrating how users rely on models for trust-heavy, consultative questions.
- ChatGPT leads in share with 40% to 60% of sessions, but rivals like Perplexity and Copilot are gaining traction across key industries.
Changing on the fly: Marketers are already adjusting. More than a third (35.2%) of US marketers plan to boost community or social presence, 25.7% are developing authoritative content for AI citations, and 14.9% are investing in digital PR, per Fractl, Search Engine Land, and MFour. Only 12.1% say they’ll make no changes.
Our take: Search engine visibility and AI discovery are the two areas brands need to focus on to drive reach. The latter is growing faster than anyone expected.
Marketers who adapt now—by tracking AI sessions, restructuring content for conversational interfaces, and optimizing across multiple models—will own the next wave of discovery. Those who don’t risk watching competitors capture visibility in AI answers while their own content fades from discovery altogether.
Treat AI as a primary discovery channel, not a test bed—doing so keeps your content visible and relevant as traditional search fragments.