Elizabeth Sramek
Published 11 October 2025
Quick Answer
Founded: 2022 by Aravind Srinivas (CEO), Denis Yarats, Johnny Ho, and Andy Konwinski — all with backgrounds at Google, Meta, and DeepMind.
Ownership: Venture-backed. Major investors include NVIDIA, Jeff Bezos (via Bezos Expeditions), IVP, NEA, and Institutional Venture Partners. The company has raised over USD 500 million at a reported USD 9 billion valuation as of 2025.
Business model: Freemium — free tier with usage limits, paid订阅 (USD 20/month for Pro). Revenue comes from subscriptions, API licensing, and enterprise deals.
How it differs from Google: Perplexity is an answer engine, not a search engine. It synthesises information from multiple sources into a direct answer with citations, rather than returning a list of links. It is designed for research tasks; Google is designed for discovery tasks.
Legal controversy: Multiple publishers have accused Perplexity of scraping content without proper attribution and potentially violating copyright. This remains an unresolved legal and ethical debate in the AI industry.
Perplexity AI has become one of the most discussed companies in the AI landscape since its founding in 2022. Often described as an “answer engine” or “AI-powered search engine,” it represents a fundamentally different approach to information retrieval than the traditional search engine model that Google normalised over two decades.
Understanding who owns Perplexity AI, how it is structured, how it makes money, and how it compares to traditional search engines is essential for anyone who wants to understand the competitive dynamics of the AI search market — and the legal and ethical questions that surround AI-powered content synthesis.
Perplexity AI was founded by four co-founders, all of whom bring significant machine learning research credentials from some of the world’s leading AI organisations:
Aravind Srinivas (CEO): Previously a research scientist at Google Brain and DeepMind, where he worked on large language models and reinforcement learning. His academic background includes a PhD from UC Berkeley (partially completed, then left to start Perplexity).
Denis Yarats: Mathematician and ML researcher who previously worked at Meta AI (FAIR), where he focused on reinforcement learning and robotics.
Johnny Ho: Former competitive programmer and researcher with experience at Quora and Meta, specialising in ML systems and ranking algorithms.
Andy Konwinski: Systems researcher from Google, where he worked on large-scale distributed computing systems. He was part of the team that developed Apache Mesos.
This founding team is unusual in its depth of genuine ML research experience. Many AI startups are founded by entrepreneurs who apply AI tools; Perplexity was founded by people who have built the underlying ML systems at some of the world’s most sophisticated AI research organisations.
Perplexity AI has raised substantial venture capital, reflecting investor belief in its market position and growth trajectory:
The company has raised over USD 500 million across multiple funding rounds. Its most recent valuation reached approximately USD 9 billion, making it one of the most valuable AI companies globally that has not yet had a liquidity event (IPO or acquisition).
Key investors include:
Perplexity AI operates on a freemium model:
Free tier: Users can ask an unlimited number of queries but with usage restrictions on the most powerful models and features. The free tier is monetised primarily through user data (anonymised behavioural patterns and query data) and brand exposure.
Perplexity Pro (USD 20/month): The paid subscription unlocks access to the most capable language models, including GPT-4o and Anthropic Claude, higher usage limits, and priority access to new features. The Pro tier targets power users — researchers, analysts, writers, and professionals who rely on AI-assisted research as a core part of their workflow.
API licensing: Perplexity has begun licensing its API to enterprise customers who want to embed its answer engine capabilities into their own products and workflows.
Enterprise deals: Custom deployments for organisations that want Perplexity’s search capabilities integrated into enterprise knowledge management systems, intranets, or proprietary data environments.
The comparison to Google is natural but misleading if taken too far. The two products serve overlapping but distinct information needs:
Search engines return links; answer engines return answers. When you search on Google, you receive a ranked list of web pages. You then have to read through those pages to find the answer to your question. Perplexity synthesises a direct answer from multiple sources and presents it immediately — with citations so you can verify the source.
Different use cases: Google is optimised for discovery — when you do not know exactly what you are looking for, want to browse multiple perspectives, or are looking for a specific website. Perplexity is optimised for research — when you have a specific question, want to understand a topic quickly, or need to synthesise information from multiple sources.
Different revenue models: Google’s revenue comes almost entirely from advertising, which creates incentives around user attention and engagement. Perplexity has no advertising in its current model, which means its incentives are aligned around answer quality, not click-through rates.
Perplexity has faced significant criticism from publishers and content creators who allege that the platform scrapes content to generate answers without providing meaningful traffic to the source websites. Major news organisations and tech publications have documented cases where Perplexity’s answers appear to closely paraphrase copyrighted content without adequate attribution.
Perplexity’s response has been to emphasise its citation practices — every answer includes links to source material — and to argue that it drives referral traffic to publishers. Critics counter that the referral traffic is minimal compared to the traffic publishers would receive if users visited their sites directly.
This controversy is part of a broader legal and ethical debate about how AI systems should interact with copyrighted content. The legal framework is still evolving, and the outcome of several high-profile lawsuits will significantly shape how AI answer engines can operate in the future.
Perplexity AI represents a credible challenge to the assumption that search is a Google monopoly. Backed by a research-grade founding team and substantial venture capital, it has built a product that genuinely serves a different information need than traditional search — one that is optimised for research tasks rather than discovery.
The company faces significant challenges: the legal uncertainty around content usage, the competitive response from Google and Microsoft, and the fundamental question of whether a subscription-only model can scale to the user base needed to sustain a USD 9 billion valuation. But its differentiated product, strong team, and clear market positioning make it one of the most interesting AI companies to watch.
By Elizabeth Sramek | Category: AI & Tools