Comet is Perplexity's AI-powered web browser, built around an agentic assistant that can read the page you are on, answer questions about it, and carry out multi-step tasks on your behalf rather than just returning a list of links. Instead of typing a query, opening ten tabs, and reading each site yourself, you ask Comet's assistant a question and it navigates, reads, compares, and summarizes across sources — presenting a synthesized answer with citations to where it looked. For law firms, the significance is not the browser itself but the shift it represents: the research layer between a prospect and your website is becoming an AI agent, not a human scrolling search results.
Comet is built on Chromium, so it behaves like a familiar browser — tabs, extensions, bookmarks — with Perplexity's answer engine woven directly into the interface as a persistent sidebar assistant. The assistant has context on whatever you are viewing, which means it can summarize a long statute page, compare two law firms' service pages side by side, or pull the key facts out of a directory listing without you leaving the tab. This is qualitatively different from opening perplexity.ai in a normal browser: the agent operates inside your browsing session, with awareness of the pages themselves.
We describe Comet's capabilities qualitatively here on purpose. Perplexity has released Comet progressively, expanding access over time, and public adoption figures shift quickly — so rather than cite numbers that may be stale, the durable point for legal marketers is directional: agentic browsing is a real, shipping product category, and the firms that understand how it reads and selects sources will be visible inside it. For the foundational mechanics of how Perplexity chooses what to cite, start with how Perplexity ranks law firms.
In the classic model, a prospect searches "personal injury lawyer near me," scans a page of blue links and map results, opens a few firm sites, and forms an impression by reading. Every one of those firm sites gets a fair chance to make its case. Agentic browsing compresses that journey: the prospect asks the assistant to "find and compare the best-reviewed personal injury firms in my area and tell me which handle motorcycle accidents," and the agent does the reading — visiting sites, extracting facts, and returning a short ranked comparison. The prospect may never manually open your homepage at all.
This changes three things for firms:
The behavioral upshot is that discovery, comparison, and shortlisting increasingly happen through an intermediary that reads far more literally and far faster than a person. The firms that win are the ones whose public information is complete, consistent, and machine-legible — the same discipline that drives Perplexity citations for law firms, now applied to an agent that acts on what it reads.
Because Comet's assistant is powered by the same retrieval-and-citation engine as Perplexity's core product, the work that makes you visible in Perplexity answers is the work that makes you visible in Comet. There is no separate "Comet optimization" — there is good AI-search hygiene, expressed through an interface that happens to act on the results. That is reassuring: firms that have already invested in citable content are positioned for Comet without new work, and firms that have not now have a sharper reason to start.
Concretely, prioritize these to be well-represented when an agent reads your site:
There is also a defensive angle. Because Comet's assistant summarizes and compares your firm, an inaccurate or thin public footprint gets amplified: a wrong practice area or a stale review profile becomes a line in an agent-generated comparison a prospect reads as fact. Auditing what the assistant says about you — and fixing the underlying sources — is the agentic-era version of Perplexity reputation management.
Comet is not only a channel that reaches your prospects — some lawyers and staff will use it themselves for research, and that raises professional obligations. An agentic browser that can read pages and act on your session is a powerful research aid for public-source work (statutes, case summaries, opposing-counsel background), but the confidentiality rules that govern any public AI tool apply fully: never route client-identifying facts, privileged material, or matter specifics through a public assistant, and understand what the tool retains before adopting it for substantive work.
Firms should also verify agent-produced research the way they verify any secondary source. An agentic summary of a limitation period or a statutory threshold is a starting point, not authority — the underlying primary source must be checked, because a confident synthesis can still misstate a jurisdiction-specific rule. Treat Comet's output as a research assistant's memo: useful, fast, and always subject to a lawyer's independent verification against the primary law.
Comet is best understood as an early, concrete instance of a broader trajectory: search that acts. The strategic response is not to chase one browser but to make your firm's public information complete, consistent, machine-legible, and genuinely useful — so that whatever agent reads it, from Comet to ChatGPT's browsing mode to Google's agentic features, represents you accurately and includes you in the shortlist. Interactive, quotable assets help disproportionately here; tools like our legal wizards answer procedural questions in a structured form agents can extract and cite, giving your firm a reason to appear that a generic services page does not.
A practical first step is to read your own firm the way an agent would: open Comet or Perplexity, ask it to compare firms in your practice area and market, and note what it says about you — what facts it gets right, what it gets wrong, and whether you make the shortlist at all. That single exercise usually surfaces a punch list of missing facts, inconsistent data, and thin pages that, once fixed, materially improves how every agent represents you.
The firms that thrive as browsing becomes agentic will be the ones whose public authority was already deep — not the ones scrambling to react. That is precisely the program LexScale.ai runs: complete, schema-rich, consistently structured content that AI agents read, trust, and act on, across Canada and the United States. To make your firm agent-ready, book a free strategy call, explore the commercial playbook at get cited in Perplexity for law firms, or work the broader Perplexity hub. For the strategy layer, see our AI SEO for law firms service.
LexScale.ai builds the content, schema, and entity architecture that gets law firms cited by Perplexity, ChatGPT, Gemini, and Google AI Overviews — across Canada and the United States.
Book a Free Strategy Call →Ready to grow your firm with AI?