Private Equity is Backing Businesses that Could Rewrite the Outsourcing Model
For most of this year the AI-and-outsourcing story has played out in Bangalore and Manila: incumbent IT-services and BPO giants repricing contracts as generative AI eats into the labour-arbitrage model that built them. Last quarter, the story picked up a second front. Anthropic announced a US$1.5bn joint venture with Blackstone, Hellman & Friedman and Goldman Sachs to build what the four call an "AI-native enterprise services firm," backed by Apollo Global Management, General Atlantic, Leonard Green & Partners, GIC, and Sequoia Capital. OpenAI did the same thing a week later, launching its own Deployment Company with $4.0bn from TPG, Advent, Brookfield, and Bain Capital. The capital that funds buyouts is now funding the machine that competes with BPO and consulting altogether.
The Deal That Changes the Maths
The target market is not small. Fortune's read on the venture is the sharp one: enterprises spend roughly six dollars on services for every dollar they spend on software, a multi-trillion-dollar pool that consulting has owned for decades and that AI-native firms are now moving to capture. The new venture embeds Anthropic engineers directly inside client companies' operations rather than sell another subscription tier. Blackstone president Jon Gray described the goal as building a company to "deploy Anthropic's incredible technology across a range of businesses in our portfolio and beyond," aimed at expanding the pool of implementation partners available to enterprise buyers. Anthropic CFO Krishna Rao put the commercial logic plainly: "Enterprise demand for Claude is significantly outpacing any single delivery model."
This is not a subtle shift for private equity. Sponsors have spent two decades buying BPO and IT-services businesses precisely because labour arbitrage was a durable, financeable cash flow. Now some of the largest names in the industry are backing the technology built to replace it. Read alongside this month's other pieces on India and the Philippines, the logic holds together: buyers of services increasingly want outcomes, not headcount, and the capital is following that preference upstream, into who delivers the outcome, not just how it's priced.
There is an important caveat. The incumbents are not standing still, and our conversations with enterprise technology companies suggest AI-native services will not simply replace offshore delivery.
The Gap Between the Pitch and the Desk
Private equity's own AI adoption inside its portfolio companies tells a more uneven story than the venture announcements suggest. FTI Consulting's 2026 Private Equity AI Radar, a survey of 200 fund and operating leaders, found 95% of AI initiatives meet or exceed their original business case. The same survey found only 36% of portfolio companies actually use AI in day-to-day operations, and just 7% describe it as fully integrated across the portfolio. Bain's global private equity research puts a similar number on production use: roughly 20% of portfolio companies have a generative AI use case live with concrete results. EY reports 84% of US private equity firms have appointed a Chief AI Officer, and two-thirds expect to steer more than a quarter of their 2026 budget toward AI.
The pattern echoes this month’s other two pieces: adoption is moving faster than organisations’ ability to measure and operationalise it. Nearly every AI programme clears its hurdle rate on paper. Very few reach the people who would use them daily. PwC's mid-2026 outlook confirms AI has moved from a nice-to-have to a formal diligence line item: sponsors are now testing how exposed a target is to AI disruption, whether it has the data foundations to adopt AI, and whether management has a credible plan, as part of the investment thesis, the value creation plan and the exit story itself.
Same Shock, Closer to Home
None of this is a story about somewhere else. Earlier this year, Australian logistics software maker WiseTech Global said it would cut about 2,000 roles, close to 30% of its 7,000-strong global workforce, over the following two years. Chief executive Zubin Appoo said, "the era of manually writing code as a core act of engineering is over." The market's response: WiseTech shares +11% on the day.
Commonwealth Bank (CBA) has been running the same play in reverse. Since early 2026, the bank's Microsoft-built AI messaging platform has been resolving close to nine in every ten customer chat conversations without a human agent. Earlier this quarter, CBA had wound back its contract with Nutun, a Johannesburg-based outsourcing provider, ending hundreds of offshore chat-support roles; the Finance Sector Union puts the bank's total 2026 cuts at close to 800. It is the ISG "AI-enabled insourcing" finding from this month's BPO piece, in miniature: a large enterprise using AI to take outsourced work back in-house, not send more of it offshore. (CBA has form here: in 2025 it reversed 45 AI-attributed job cuts after the union showed call volumes had actually risen, the same attribution question this month’s other pieces raise about BPO providers).
What This Means, by Seat: North Ridge Partners' View
If you're a tech company deciding whether to outsource:
The choice is no longer simply BPO provider versus build in-house. AI-native services firms are creating a third model, while conventional offshore providers are rapidly adding AI to their own delivery stacks. Buyers should resist locking themselves into long-duration arrangements whose economics depend entirely on FTE growth. At the same time, they shouldn’t leap blindly into outcome pricing. The measurement systems required to attribute outcomes reliably often need to be installed first, so a sensible route is a contractual glide path: establish the baseline, measure productivity, progressively share the gains, and migrate toward outcome-based economics as the data becomes trustworthy.
If you're private equity buying tech companies:
AI exposure is now a formal diligence line item, not a bolt-on. Treat an AI-driven value-creation story with the same scepticism this month's other pieces apply to BPO providers' own numbers: ask who owns the attribution, and what share of the workforce actually uses the tool daily, not just whether the pilot cleared its business case.
If you're a venture firm, funding software startups:
The services layer that PE-backed AI-native firms are now targeting is roughly six times the size of the software layer venture capital traditionally funds. That is a direct threat to any startup whose moat depends on professional-services or integration revenue sitting on top of the product. Push portfolio companies to build measurement and attribution into the product from day one. A start-up that can prove ROI cleanly will out-compete one with a marginally better model, because provable value, not model quality, is what buyers are now demanding before they will pay outcome-based pricing.
Read together with this month's pieces on India and the Philippines, the destination nevertheless looks remarkably consistent: fewer people attached to each unit of output, more automation inside delivery, and progressively greater pressure to price what gets achieved rather than how many hours were spent achieving it. The winners may be AI-native challengers, reinvented incumbents or internal enterprise teams. Most likely, it’ll be their customers!