Scale AI to cut 14% of staff after Meta investment, CEO departure
San Francisco – Scale AI is laying off hundreds of employees from its data-labelling business, one month after Meta Platforms invested US$14.3 billion (S$18.3 billion) in the start-up and hired away its chief executive officer.
The company cut 200 full-time employees, about 14 per cent of its global workforce, and will provide severance, Scale spokesperson Joe Osborne said on July 16. Scale will also stop working with 500 of its thousands of global contractors, he said.
The move is aimed at 'streamlining our data business to help us move faster,' Mr Osborne said, adding that Scale plans to staff up in other areas including enterprise and government sales.
In a note sent to Scale employees on July 16, interim CEO Jason Droege said the layoffs were a result of the data labelling business bringing in too many people too quickly over the last year. That led to 'too many layers, excessive bureaucracy, and unhelpful confusion about the team's mission,' Mr Droege wrote in the memo, which was viewed by Bloomberg News.
Mr Droege added that 'shifts in market demand' also contributed to the decision to restructure. Following the deal with Meta, some of Scale's most prominent customers have phased out work with the start-up – including OpenAI and Alphabet's Google – according to reports from Bloomberg and others.
Founded in 2016, Scale has long been the best-known name in the market for helping tech firms label and annotate the data needed to build AI models. It generated about US$870 million in revenue in 2024 and expects US$2 billion in revenue in 2025, Bloomberg News reported in April.
In June, Meta finalised its multibillion-dollar investment in Scale, taking a 49 per cent stake in the company. As part of the deal, co-founder Alexandr Wang left the start-up to lead a new superintelligence unit at Meta, part of the Facebook parent company's multibillion-dollar investment to catch up on AI development.
Despite its position as a leader in the market for providing a key ingredient needed for building AI models, Scale faces a growing raft of rivals including Turing, Invisible Technologies, Labelbox, and Uber Technologies, all of which also offer services to meet AI developers' bottomless need for data. As some of Scale's clients worried about Meta getting added visibility into their AI development process, competing services have said they've seen a surge in interest from customers.
Scale is one of several AI companies that have seen key talent hired away in the last year, without being acquired. Most recently, Google inked a US$2.4 billion deal with AI coding start-up Windsurf, hiring its CEO and several of its top staffers. The deals raise the question of what happens to the employees left behind after the CEO joins a larger firm. In Windsurf's case, the start-up was quickly bought by another AI company. At Scale, the company plans to leverage its war chest, Mr Droege said.
'We remain a well-resourced, well-funded company, and today's announcement will allow us to accelerate new investments and add resources where necessary,' he wrote in the memo. Mr Droege said the start-up plans to hone the scope of its data-labeling business to focus on projects related to coding, languages and audio.
Mr Osborne said the company plans to hire hundreds of people in the second half of 2025 for efforts including making custom AI applications and working with government agencies in the United States – such as the Defense Department – and around the world. Those parts of business are currently making nine figures in revenue, Mr Droege previously told Bloomberg. BLOOMBERG

Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles


CNA
2 minutes ago
- CNA
Cisco forecasts higher-than-expected quarterly revenue on increased demand
Cisco Systems forecast first-quarter revenue above Wall Street estimates on Wednesday, as the artificial intelligence boom boosted demand for its networking equipment from cloud customers. Enterprises such as Cisco have been benefiting as more businesses move workloads to cloud environments and modernize campus footprints to prepare for AI technologies. "The AI infrastructure orders we received from webscale customers in fiscal 2025 were more than double our original target, indicating a massive opportunity ahead as we lead the required architectural shift and build the critical infrastructure needed for the AI era," CEO Chuck Robbins said. The company expects its revenue to be between $14.65 billion and $14.85 billion during the first quarter, compared with analysts' average estimate of $14.62 billion, according to data compiled by LSEG.


CNA
2 minutes ago
- CNA
Cisco forecasts first-quarter revenue above estimates
Cisco Systems forecast first-quarter revenue above Wall Street estimates on Wednesday, as the artificial intelligence boom boosted demand for its networking equipment from cloud customers. The company expects its revenue to be between $14.65 billion and $14.85 billion for the quarter, compared with analysts' average estimate of $14.62 billion, according to data compiled by LSEG.


CNA
6 hours ago
- CNA
Tencent says US AI chip import situation unclear as governments negotiate
BEIJING :Chinese tech giant Tencent said on Wednesday it lacks clarity on U.S. AI chip imports as Beijing and Washington continue negotiations, though the company maintains it has sufficient inventory for its AI operations. "We don't really have a definitive answer on the import situation yet. There are a lot of discussions between the two governments. We are waiting to see what exactly comes out of that," Tencent President Martin Lau said during a post-earnings call. The comments come amid ongoing tensions over advanced semiconductor trade between the world's two largest economies. Nvidia's H20 chips, designed specifically for the Chinese market, have been at the center of recent regulatory scrutiny from both governments over security concerns. Lau said the uncertainty would not constrain Tencent's AI ambitions, noting the company has adequate chip supplies for AI model training and multiple deployment options for AI services. The Shenzhen-based company reported strong second-quarter results, with revenue rising 15 per cent to 184.5 billion yuan ($25.7 billion), beating analyst estimates of 178.5 billion yuan according to LSEG data. Gaming remained a key growth driver, with domestic revenue rising 17 per cent to 40.4 billion yuan and international revenue climbing 35 per cent to 18.8 billion yuan. Marketing services revenue increased 20 per cent year-on-year to 35.8 billion yuan, bolstered by the use of AI to enhance the targeting of adverts. Net profit for the quarter reached 55.6 billion yuan, surpassing analyst expectations of 52.3 billion yuan, LSEG data showed. Tencent has accelerated AI investments over the past two years, though capital expenditure has moderated recently. After spending 36.6 billion yuan in the fourth quarter of 2024 and 27.5 billion yuan in the first quarter, its capex fell to 19.1 billion yuan in the second quarter. Lau signaled a more measured approach going forward as the company seeks sustainable monetization from its AI initiatives and to "spend smartly". The company has developed its own large language model, Hunyuan, launching the latest "Turbo S" version in February. But Tencent has also embraced third-party models, notably integrating DeepSeek's technology across its platforms including WeChat, which has over 1 billion monthly active users. ($1 = 7.1756 Chinese yuan renminbi)