How AI Coding Companies Advertise — A Competitive Breakdown

We ran competitive ad audits on three of the AI coding category's most visible players. Read together, they reveal a strange asymmetry: the most successful brand in the category is the only one actually selling an AI coding tool by name.

AI coding assistants are the fastest-spreading developer technology in history. Stack Overflow's 2025 Developer Survey found 76% of developers using or planning to use AI tools in their workflow; JetBrains' State of Developer Ecosystem found 49% adoption of AI coding assistants in production code. Using ULUK, we ran competitive audits on three of the category's most visible players — Cursor, GitHub Copilot, and Claude Code (Anthropic). Each audit analyses a 30-day window of in-market campaign activity across 42 strategic dimensions. Read side by side, the three reveal something unusual: in a category where the buyer is shopping for an AI coding tool, only one of the three advertisers is actually advertising one. Here is what their advertising shows.

1. Cursor is the only advertiser branding itself as an AI coding tool

The clearest pattern across the three audits is who names the product and who doesn't. Cursor names Cursor in every active creative. GitHub never names Copilot — its in-market advertising leans entirely on GitHub-the-platform with an "agentic AI + DevOps" rotation. Anthropic positions Claude Code as one feature pillar among many, alongside Claude Opus 4.8, the Microsoft Foundry partnership, and Claude Cowork. In a category where every buyer's first search is some version of "the best AI coding tool," Cursor is the only brand actually putting its name in front of those words. The asymmetry is wider than any other dimension we measured in the study.

2. Cursor runs the most mature enterprise B2B playbook in the category

For a product that grew primarily through developer Twitter and GitHub stars, Cursor's paid advertising is remarkably sophisticated. Across just a handful of creative variants, Cursor stacks every classic enterprise B2B move:

• Account-level ABM personalization. MESSAGE-format DMs signed by "Emily," targeting engineering leaders by first name and naming six current customers (NVIDIA, Stripe, Salesforce, Adobe, Hilton, Fox). No other advertiser in the category runs ABM.
• Third-party analyst validation. Gartner Magic Quadrant Leader recognition surfaced across multiple creative variants. Only advertiser with this proof texture.
• Owned original research. The Cursor Developer Habits Report — "the most comprehensive dataset on AI coding in the world." Only owned research engine in the category right now.
• Direct-incentive conversion offer. $500 Teams credit to move teams to the Cursor Teams plan. Only direct incentive in the category.
• Quantified customer outcomes. "Rippling moved from 150 to 500+ engineers on Cursor" appears in active creative. Specific, measurable, social-proof-dense.

The playbook reads like enterprise B2B SaaS circa 2010 — which is exactly what the category is becoming.

3. The AI coding category is in the middle of a paid-media transition

Most successful AI coding tools grew without paid media. Cursor itself crossed $500M ARR primarily through developer Twitter, GitHub stars, and word-of-mouth. The shift to paid-social advertising in late 2025 and early 2026 is therefore a signal in itself: the category is transitioning from individual-contributor adoption to enterprise procurement. JetBrains' 2025 data confirms it — enterprise engineering leadership now drives 64% of AI-coding tool buying decisions in large organisations, up from 41% a year earlier. The buyer has moved up, and the marketing has to follow. Cursor has made that shift; GitHub is running enterprise IT positioning that doesn't even name the product; Claude Code's creative is still feature-led and aimed at the individual developer. The category's marketing playbook is being written in 2026, and the advertiser that establishes the dominant playbook now sets defaults for years.

4. GitHub's Copilot brand is quietly being subsumed

The most consequential finding for the rest of the category is what's happening to Copilot specifically. GitHub's in-market creative no longer mentions Copilot by name in any 30-day-window ad. Every variant positions "GitHub" (the platform) around "agentic AI" + "DevOps" — broad enterprise IT framing rather than product-specific marketing. Whether this is intentional consolidation under Microsoft's broader agentic platform narrative or an unintended consequence of marketing reorganisation, the effect is the same: the Copilot brand equity is being diluted. For Cursor and Claude Code, this is uncontested category-naming whitespace. Every paid impression that names "AI coding tool" goes to whichever brand puts itself in front of the words. Right now, only one brand is doing that.

Why this matters

None of the above is opinion. It is a structured read of what these companies actually put in-market, every finding tied to a real campaign. That is what a ULUK competitive audit produces — and the full synthesis behind this piece is public.

This analysis is based on a trailing 30-day window of in-market campaign activity in the United States, drawn from a representative sample of each company's library. Findings reflect creative, messaging and format patterns observed in that sample. Industry-trend references are sourced from Gartner's Magic Quadrant for AI Code Assistants, the 2025 Stack Overflow Developer Survey, JetBrains' State of Developer Ecosystem 2025, and Forrester's State Of Business Buying 2026, each cited in the underlying Cursor Competitive Strategy Synthesis.