OpenAI is establishing an “ads integrity” team to ensure the responsible introduction of advertising within ChatGPT. This initiative aims to prevent fraudulent ads and maintain user trust as the company prepares for a small test of ads in its free version and Go tier in the United States, with a requirement for advertisers to commit a minimum spend of $200,000. The new team will focus on developing systems to verify advertisers’ identities and manage risks associated with scam ads, a significant concern for major tech companies with self-service ad platforms. According to a recent investigation by Reuters, Meta removed over 134 million scam ads in 2025, highlighting the challenges in maintaining ad integrity.
OpenAI is set to hire hundreds of technical consultants to bolster its consulting team, aimed at assisting large corporations in developing custom artificial intelligence applications. This strategic move comes as OpenAI is preparing to launch a new enterprise offering designed to streamline businesses’ AI efforts and improve customer service capabilities. Currently employing around 60 engineers dedicated to customizing AI models for various clients, OpenAI aims to increase the share of revenue from business customers from approximately 40% to 50% by the end of the year, translating to an estimated $15 billion in revenue for 2026. The company has been in informal discussions regarding a potential initial public offering, indicating significant growth and market interest in its services.
Why do we care?
These look like two separate stories, but they’re the same one: OpenAI is moving from platform provider to services competitor, and ads are how it funds the transition.
OpenAI describes the ads integrity team as balancing “rapid revenue growth and responsible platform stewardship.” That’s not alignment—that’s acknowledged tension. We’ve seen how this plays out. Internal documents reported by Reuters showed Meta projected billions in scam-ad revenue while users were flooded with high-risk ads. Meta has integrity teams. It has fraud detection. It still makes billions from scams because the economic incentive overwhelms enforcement at scale.
OpenAI’s $200K minimum advertiser spend is a temporary moat. It keeps out low-end scam volume—for now. That’s a launch constraint, not a durable control. When that floor drops—and it will, because growth demands it—the same dynamics apply.
Now connect that to OpenAI hiring hundreds of forward-deployed engineers to embed inside enterprise clients. This mirrors the Palantir playbook: high-touch implementation that creates dependency rather than durable internal capability. When those engineers leave, the client’s AI implementation becomes orphaned.
Here’s what most MSPs are missing: OpenAI isn’t just selling to enterprises. Through Thrive Holdings, it has taken ownership stakes and embedded engineers inside operating companies—including Shield, an MSP that has now received a second $100 million investment. Shield reaching $100 million ARR with embedded OpenAI engineers isn’t organic market success—it’s a controlled experiment validating OpenAI’s services model ahead of an IPO.
The mistake MSPs are making is reading “OpenAI is hiring consultants” as ecosystem investment instead of ecosystem disintermediation. They keep building undifferentiated practices on OpenAI APIs. Meanwhile, Shield—with embedded OpenAI engineers and $200 million in backing—is competing for the same clients with superior access to the underlying platform.
This isn’t about ads or consultants in isolation. It’s OpenAI vertically integrating into services delivery—ads fund it, consultants operationalize it, and Shield proves it.
MSPs who don’t see this coming will realize they built on a platform that now competes with them—and has better access to its own technology than they ever will.