Meta’s 2025 investment in Scale AI reportedly put the data company’s relationship with Google—its largest customer—under immediate pressure. Reuters reported that Google planned to move most of its work elsewhere, putting about $200 million in expected 2025 spending at risk. That was not confirmation that Google canceled a single $200 million contract or ended every engagement.
What Meta’s Scale AI investment involved
In June 2025, Meta invested about $14.3 billion for a 49% stake in Scale AI, according to Bloomberg’s report on the transaction. Scale was valued at more than $29 billion including the investment. Scale founder Alexandr Wang left to join Meta’s AI organization.
This was a large minority investment, not a purchase of all of Scale. The structure gave Meta substantial economic exposure and a closer connection to Scale’s capabilities while leaving Scale operating as a separate company. It also created a perception problem: Scale supplied services to companies competing with Meta in AI.
What Google reportedly planned—and what the $200 million meant
On June 13, 2025, Reuters reported that Google planned to cut ties with Scale and was discussing alternatives. The report, based on people familiar with the matter, described Google as Scale’s largest customer and put its planned 2025 spending on Scale’s human-labeled training data at about $200 million. Reuters’ report, reproduced by Investing.com, did not establish a publicly announced termination of every contract.
The distinction matters for understanding the financial impact. Planned annual spending is not necessarily a signed minimum commitment, recognized revenue, or a cancellation fee. The report supports saying that roughly $200 million in expected business was at risk—not that Google definitively canceled a $200 million contract or that Scale lost that amount in revenue.
Why a data supplier’s neutrality matters
Scale is more than a basic labeling contractor. Its work can include preparing and curating data, human annotation, expert judgments, model-output grading, safety evaluation, and testing AI systems. Such services help developers assess and improve models, and may involve work tailored to a customer’s methods and priorities. The Associated Press described Scale’s role and its customer base across AI and other industries in its coverage of the investment.
For a customer such as Google, the concern is not proof that Meta received access to confidential material. The available reporting does not establish that any customer data or secrets were transferred. Rather, a direct competitor’s ownership stake and the founder’s move to Meta could make a shared supplier seem less independent, even if contracts, access controls, and internal firewalls were in place.
Customers may worry that a vendor’s staff could encounter sensitive research priorities, evaluation approaches, or model-development needs while serving multiple AI labs. The more specialized the human work, the more consequential trust and separation can become. A customer can therefore move sensitive projects elsewhere without alleging a breach.
PC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Crashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minuteHow the trade-off looked for each company
- Meta: The investment offered closer access to data, evaluation expertise, and Scale’s founder. But if rival customers reduced their work, Meta’s investment could weaken the independent supplier it sought to align with.
- Google: Moving work could reduce perceived exposure to a competitor-backed vendor and improve negotiating options. The trade-off is the time and expense of qualifying replacements, transferring workflows, and maintaining consistent quality.
- Scale: The investment brought substantial capital and a closer relationship with Meta, but risked making Scale less acceptable as a neutral provider to competitors.
- Alternative suppliers: Google’s reported search for other vendors created an opportunity for competitors to win work, though the reporting does not identify a confirmed replacement for all of Scale’s services.
OpenAI’s reduction in Scale work is a separate case
Bloomberg reported on June 18, 2025, that OpenAI was phasing out work with Scale after Meta’s investment. OpenAI said it had already been reducing its reliance on Scale before the transaction and that Scale accounted for only a small share of its overall data needs. Bloomberg’s account therefore does not support attributing OpenAI’s entire change to Meta’s deal.
Why the deal drew scrutiny beyond customer relations
The arrangement combined a major investment with the recruitment of Scale’s founder, prompting lawmakers to examine similar structures under the label “reverse acquihires.” A Senate letter cited the Meta–Scale transaction in a broader inquiry into Big Tech investment and talent arrangements. The letter reflects questions raised by lawmakers; it is not a finding that the transaction was unlawful.
What remains unclear about Scale’s financial fallout
The reporting cited here does not establish whether Google completed a full migration from Scale, how much work it ultimately moved, or how much revenue Scale lost. It also does not show whether Scale replaced any reduced business, whether Meta’s stake gave it access to customer-confidential information, or whether Google later resumed or retained particular projects. A reported reduction in a customer relationship is not enough to calculate the supplier’s net financial impact.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Scale’s 2026 leadership change
On July 30, 2026, Axios reported that former Google Cloud COO Francis deSouza had become Scale’s chief executive, replacing interim CEO Jason Droege. Axios’s report documents a leadership change, not a restoration of Google’s commercial relationship or proof that the earlier customer concerns were resolved.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
What companies should assess when replacing an AI-data vendor
Switching suppliers is not simply a matter of transferring a labeling queue. A procurement team evaluating alternatives should compare the protections and operating capabilities that matter for its particular data and models:
Quick Recap
- Confidentiality and separation: Ask how customer projects, personnel, systems, and data are separated—especially when the vendor serves direct competitors.
- Workforce and location: Establish who performs the work, where they are located, how they are vetted, and whether subcontractors are involved.
- Quality control: Review expert qualifications, adjudication procedures, error measurement, and auditability for the task at hand.
- Security and accountability: Check data handling, retention and deletion terms, incident procedures, and contractual remedies for quality failures or unauthorized disclosure.
- Continuity and capacity: Confirm that the provider can handle the required volume and expertise, and plan for migration, calibration, and quality checks before moving sensitive work.
- Commercial terms: Distinguish forecast spending from committed minimums, and understand pricing, scope, turnaround, and change-control terms before treating a budget estimate as a contract value.
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
