Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Repair Windows errors before they cause bigger problems3Fix the driver behind crashes, sound loss and screen glitchesSome links on this page are affiliate links: if you buy through them we may earn a commission, at no extra cost to you.
The “very dark AI company” in this headline is Scale AI, a provider of data-labeling and model-evaluation services that also has defense contracts. Meta did not buy it outright: reporting put Meta’s June 2025 investment at about $14.3 billion for a 49% non-voting stake. The controversy behind the “dark” label centers on military applications and allegations about the treatment of data workers, not an official finding that Scale acted illegally.
What Meta invested in
Scale AI is an AI data and infrastructure-services company, not a consumer chatbot maker. Founded by Alexandr Wang and Lucy Guo, it prepares and labels data and helps customers evaluate and test machine-learning systems. That work can include human review of data or model responses, which helps developers assess and improve their systems. Scale describes its services at scale.com; the Associated Press also outlined the company and Meta deal in its coverage.
Data preparation and evaluation are less visible than chips or chatbot demonstrations, but they are part of the infrastructure behind AI development. Scale’s position across the industry also means the Meta deal has implications beyond the two companies: Scale has served multiple AI developers, including companies competing with Meta.
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 & 11Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchWhat the $14.3 billion deal did—and did not—buy
Scale announced the transaction on June 12, 2025, saying it had received a significant investment and that its valuation exceeded $29 billion. The approximately $14.3 billion amount and roughly 49% ownership figure were reported by Bloomberg and TechCrunch. Meta’s SEC filing confirms the investment was a non-voting minority interest, rather than a purchase of the entire company.
#1 Best Overall
| Deal detail | What is established |
|---|---|
| Reported investment | Approximately $14.3 billion, according to Bloomberg and TechCrunch; Scale’s announcement did not state this amount. |
| Reported stake | About 49%, described in transaction reporting; Meta’s filing confirms the interest is non-voting and minority. |
| Valuation | More than $29 billion, stated by Scale in its announcement. |
| Corporate status | Scale said it would remain independent. Meta’s SEC filing describes a minority investment, not full ownership. |
The distinction matters. A 49% economic stake is not the same as control of a company, and it does not establish that Meta receives Scale customers’ data or owns their data sets. Scale said it would remain independent and protect customer data in a statement about the deal; that is the company’s assurance, not independent verification. Meta’s filing is available here, and Scale’s customer statement is here.
Why Alexandr Wang moved to Meta
Wang co-founded Scale and had been its CEO. Under the announced arrangement, he left that executive role to join Meta’s AI effort while remaining on Scale’s board. Scale named Chief Strategy Officer Jason Droege interim CEO. Meta CEO Mark Zuckerberg later identified Wang as leading Meta’s overall superintelligence team; Nat Friedman was leading AI products and applied research, with Shengjia Zhao as chief scientist. Meta’s statement appeared in its Q2 2025 results post.
Zuckerberg’s term “superintelligence” describes Meta’s ambition, not a demonstrated capability or a claim that such systems have been achieved. Wang’s recruitment links Meta’s push to a company that has built expertise in data generation, evaluation, and government-facing work.
Rank #2
Why Meta would make such a large minority investment
The public rationale is a combination of talent, infrastructure, commercial cooperation, and strategic positioning—not simply a purchase of data. Scale said the deal would substantially expand its commercial relationship with Meta. Reporting also described the investment as part of Meta’s effort to strengthen its position against other frontier AI developers. Wang’s experience with government customers may be relevant as Meta pursues more public-sector work, though the public deal materials do not establish that this was the sole or primary reason for the investment. Scale’s announcement is here; Axios discussed the competitive and customer implications here.
- Talent: Meta brought Wang into its AI leadership effort.
- Data and evaluation capabilities: Scale’s work addresses data preparation, human feedback, and model testing that AI developers need.
- Commercial relationship: Scale said its business relationship with Meta would expand substantially.
- Competitive position: A close relationship with a major AI-services provider may matter in a race involving Meta, OpenAI, Google, Anthropic, and others.
None of these points means the investment guarantees a breakthrough. Buying access to capabilities and recruiting an experienced executive can strengthen a strategy; neither proves that Meta will achieve its stated AI ambitions.
What “dark” refers to: defense contracts, not proof of weapons control
The “dark” description is an editorial framing used by Futurism, not a formal designation. One reason for it is Scale’s work with defense agencies. Scale has described Thunderforge, a program awarded through the Defense Innovation Unit, as an AI effort intended to support military decision-making and operations. It has also announced a $500 million Pentagon AI partnership expansion centered on Scale Donovan and related defense capabilities. The company’s announcements are here and here.
Those facts warrant scrutiny, but they do not establish that Scale independently chooses targets, controls lethal weapons, or makes final targeting decisions. Three different functions should not be collapsed into one:
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
- Data labeling and evaluation: preparing or reviewing information used to train, test, or assess AI systems.
- Decision support: tools that can help people analyze information, plan, or make operational decisions.
- Autonomous weapons control: systems that select and engage targets without a person making the final decision.
The cited Scale materials establish defense contracts and military applications, including planning and decision support. They do not, on their own, show that Scale operates autonomous weapons. Wang’s congressional testimony and an earlier House hearing provide additional context on the company’s defense work: 2025 testimony and 2024 hearing material.
What is alleged about Scale’s human workers
AI data services can depend on people doing tasks such as classifying images, rating text, and evaluating model responses. That makes the working conditions and protections of annotators a relevant part of the AI supply chain, especially when the material is sensitive or disturbing. Questions include who performs the work, where they are based, how they are paid, and what safeguards protect workers and customer information.
Rank #4
A California complaint involving people performing generative-AI data-labeling work alleges labor-law violations. A complaint records claims made in litigation; it is not itself a court finding that those claims are true. The complaint is available here. Futurism’s coverage uses more forceful language, but that characterization should be distinguished from what a complaint or final legal finding establishes.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.The risks for Meta, Scale, and their customers
A large minority investment can create strategic influence and commercial ties without giving the investor the ordinary control of a subsidiary. That structure leaves several practical questions open:
The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →- Customer confidence: Scale’s customers may question whether a major investor that competes in AI can be kept separate from their confidential work. The investment does not itself prove access to customer data, but maintaining clear protections matters.
- Governance: Meta has a substantial economic interest while Scale says it remains independent. That makes the boundaries between investment, cooperation, and control important.
- Competition scrutiny: A large stake combined with the recruitment of Scale’s founder may attract questions about whether the arrangement could affect competition, even though it is not a full acquisition.
- Valuation and execution: A reported $14.3 billion is a high-stakes bet on strategic value. Scale’s capabilities and Wang’s leadership do not guarantee that Meta will produce a leading model.
- Ethics and reputation: Defense applications and allegations about labor conditions can create public, regulatory, and customer concerns.
- Concentration risk: Scale’s role serving different AI companies may become harder to sustain if customers decide they would rather use providers without a major investment from a competitor.
These are risks and questions, not proof that the deal has already harmed customers, violated competition law, or produced a model breakthrough. The available public statements establish Scale’s declared independence and its customer-data commitment, but they do not provide an independent audit of how those protections operate.
Best Value
Why the headline is both striking and incomplete
Calling Scale a “dark AI company” compresses several distinct concerns into one phrase. The phrase points readers toward the military uses of AI and the human labor behind AI data services, but it can also imply wrongdoing that the available evidence does not establish. The clearer description is that Scale sits at the intersection of AI data infrastructure, defense-related decision-support work, and a labor model that has faced legal allegations.
For Meta, the deal is a large strategic investment and a talent move, not an outright acquisition. Its significance lies in the way it connects Meta’s AI ambitions to a services provider with customers across the industry, defense contracts, and a workforce that helps prepare and evaluate AI systems.
Quick Recap
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.
