Anthropic, Goldman Sachs and Blackstone Just Launched a $1.5 Billion Company to Gut the Consulting Industry: What This Means for Agencies
On May 4, 2026, Anthropic launched a $1.5B enterprise AI services venture with Goldman Sachs, Blackstone, and Hellman & Friedman. Fortune called it "Anthropic taking shot at the consulting industry." The agency model is the explicit target.

On Sunday, May 4, 2026, Anthropic announced a new enterprise AI services company in partnership with Goldman Sachs, Blackstone, and Hellman & Friedman. The capital structure: Anthropic, Blackstone, and Hellman & Friedman each committed approximately $300 million; Goldman Sachs added approximately $150 million. Additional investors include General Atlantic, Apollo, GIC, Leonard Green, and Sequoia. Total funding: approximately $1.5 billion.
Fortune's headline on Monday, May 5, summarised the strategic intent: "Anthropic takes shot at the consulting industry."
The model is direct. Embed Claude-armed engineers inside companies to redesign workflows: the same playbook that McKinsey, Accenture, and BCG charge multi-million-dollar engagements to deliver. The pricing point and distribution channel are different. The capability target is similar.
What Most Coverage Got Wrong
The dominant framing of the announcement has been "Anthropic enters consulting." That framing is technically accurate but it materially understates what the venture is positioned to do.
If the new company were aimed at the traditional Big Three or Big Four consulting market, Fortune 500 enterprises with eight-figure consulting budgets, it would be one entrant among dozens. Those clients run formal RFP processes, have established relationships with multiple consulting firms, and have procurement teams whose job is to extract value from competitive bidding.
The Anthropic venture is not built for that market.
It is built for the mid-market PE-portfolio company. Goldman Sachs, Blackstone, and Hellman & Friedman together own hundreds of mid-market companies across their respective portfolios. Those companies have operational improvement requirements but limited internal capacity to identify and execute them. The classical solution has been to bring in outside consultants or specialised agencies on project bases.
Under the new venture, the PE owners now have direct distribution into their own portfolio companies for an embedded AI engineering team. The framing to portfolio CEOs will be straightforward: use the embedded Claude team first; if the team cannot deliver the outcome, then approve outside spend.
That is the procurement filter that most agencies and consultancies have never had to navigate before. It does not look like a competitor. It looks like a precondition.
The Implications for Sydney Agencies
The Anthropic venture's direct competitive impact on Sydney agencies depends on what the agency sells. The categories most exposed:
Process consulting and operational design. Many recruitment and marketing agencies have added these line items to their service mix in the last three years as standalone retainers came under pressure. The pitch was that the agency understands the client's domain well enough to redesign internal workflows. The Anthropic venture is offering exactly this capability to PE-owned mid-market companies, with deeper technical execution and cheaper unit economics.
AI strategy and AI implementation support. Agencies that have positioned themselves as "AI-first" partners or that bill for AI deployment guidance are in direct competitive overlap with the embedded Claude team model. The Anthropic team brings access to the most capable model, the engineering depth to integrate it, and a distribution channel that bypasses agency procurement entirely.
Fractional CMO / fractional CRO / fractional CTO arrangements. Embedded executive support is structurally similar to embedded AI engineering, and the cost comparison is unfavourable. A fractional executive at $15-30k per month versus an embedded Claude team funded centrally by the PE owner is not a cost-comparable choice.
The categories less exposed:
Brand strategy and creative direction at the senior level. The judgment-intensive, taste-driven work of defining what a brand should be at a strategic level is not what an embedded engineering team is built to deliver. Anthropic's new venture is positioned for operational redesign, not brand definition.
Specialist recruitment with deep candidate networks. The relationship-intensive, network-dependent work of placing senior or specialised talent is not addressed by the embedded Claude team model. The team can analyse data and redesign processes, but it does not bring the candidate relationships that recruitment agencies have built over years.
Client relationship management for established accounts. The work of maintaining trust and continuity with senior client stakeholders over multi-year relationships is human-dependent and difficult to disrupt with embedded AI tooling.
The Sales Cycle Implication
The most subtle effect of the Anthropic venture is on the sales cycle for outside vendors selling to PE-owned mid-market companies.
The pre-Anthropic-venture sales cycle for an agency selling into a PE-owned company looked like this: agency identifies opportunity, builds relationship with client stakeholder, develops proposal, navigates procurement, secures approval, begins engagement. The cycle typically takes 60-120 days from first conversation to signed contract.
The post-Anthropic-venture sales cycle includes a new step. The PE owner has installed a default option: the embedded Claude team. Before any outside vendor proposal can be approved, the implicit question is: has the embedded team tried this and failed? If not, the proposal gets paused while the embedded team takes a turn.
That extends the sales cycle from 60-120 days to 60-120 days + however long the embedded team needs to determine the work is genuinely beyond their capability. For agencies whose cash flow depends on sales cycle predictability, this is a structural change to forecast accuracy.
What Agencies Should Actually Do
The first step is honest diagnosis: which of the agency's current service lines sit inside the embedded-Claude-team capability boundary, and which sit outside it?
Service lines inside the boundary, process redesign, basic AI deployment, content production at scale, standard analytics, are exposed. The agency is increasingly competing not on capability but on speed, integration with existing client systems, and willingness to handle work that the embedded team finds uneconomic.
Service lines outside the boundary, strategic brand positioning, senior creative direction, specialised candidate networks, long-term relationship continuity, judgment-intensive client advisory, are protected, at least for now. These are the line items that should be priced explicitly and protected commercially.
The second step is to make the boundary visible to clients. Most agencies bundle their pricing in ways that obscure which parts of the service are in the embedded-team's capability range and which are not. Clients who do not see the distinction will assume the entire bundle is replaceable, especially when their PE owner is suggesting they try the embedded team first.
The third step is to position the agency as the natural escalation when the embedded team hits its limits. The Anthropic venture will not solve every problem at every PE-owned company. There will be situations where the embedded team determines the work requires capabilities outside its scope. The agencies that have built credibility for handling exactly those situations will be the ones the embedded team recommends bringing in.
Anthropic, Goldman Sachs, and Blackstone just installed a procurement filter at the front door of every PE-owned company in their portfolios. The agencies that walk through that filter successfully will be the ones that have made their non-replaceable value explicit, in writing, before the conversation arrives.
