OpenAI adjusts funding goals as rivals move in on market share

OpenAI's shifting funding landscape

OpenAI is currently navigating a complex private funding round as it prepares for a potential mega-AI IPO early next year. The company originally aimed to raise $30 billion at a $1.4 trillion valuation, but recent reports indicate it has had to adjust its expectations. Per the Financial Times, OpenAI slashed its funding target by $10 billion to approximately $20 billion as tech anxieties mount.

This adjustment comes during a period of significant market volatility following disappointing results from major tech firms like Google, Amazon, and Intel. Investors are currently weighing OpenAI's progress against its primary rival, Anthropic, which is nearing its own IPO. Anthropic recently reported a revenue jump to $11.6 billion in the second quarter, while OpenAI's revenue grew about 18 percent to $6.7 billion.

Despite the lowered target, the remaining funding goal is still considered a massive sum compared to previous tech waves. A group of investors from the UAE, including MGX, is expected to lead the round with participation from major firms like SoftBank and Sequoia. The situation remains fluid as the company balances these private commitments with the regulatory and permitting hurdles facing massive data center construction.

DeepSeek challenges coding leaders

China's DeepSeek is making a direct move to challenge Anthropic's leadership in the AI coding space. The company recently released Super-Agent, an open-source coding agent prototype designed to compete with Claude Code. Unlike some static tools, Super-Agent is designed to be proactive by watching developers' group chats to spot coding problems and suggest fixes automatically.

DeepSeek is well-positioned to pursue this goal due to its recent massive capital injection. The company raised about $15 billion in one of China's largest funding rounds ever, following a previous $7.4 billion raise in June. These funds are intended to support the development of software around its models as the company prepares for its own IPO.

This move highlights the growing competition between Chinese and American AI firms. While Anthropic and OpenAI remain the current front-runners, DeepSeek is joining other Chinese entities like Alibaba in the race for large-scale models. Alibaba is currently on track for potential models featuring between 5 to 10 trillion parameters, signaling a broader push for market dominance.

Samsung's push for memory share

In the hardware sector, Samsung is aggressively attempting to seize market share from SK Hynix, the current leader in AI memory. SK Hynix recently reported a sevenfold jump in operating profit and committed $45 billion to shareholders, totaling $75 billion for the year. In response, Samsung is increasing its spending by more than 60 percent this year to win back share in the high-bandwidth memory market.

The competition is centered on supply capacity for the AI boom. SK Hynix is currently sold out on its high-bandwidth memory and other memory products through 2029. Samsung, meanwhile, has reported being sold out through 2027, indicating a high demand for the specialized chips required to power large-scale AI models.

While Samsung's recent share price faced some pressure after it delayed a shareholder reward plan, its spending hike shows a commitment to the AI infrastructure race. The memory market is becoming increasingly crowded as other players like China's CXMT and YMTC emerge as numbers four and five in the rankings. This environment is being described as a period of intense competition for the memory 'king'.

Meta enters the physical AI race

Meta is expanding its reach beyond software models into the realm of physical AI by developing humanoid robots. The company built a humanoid prototype called Max in just three months by partnering with the Chinese startup Fourier. Meta views these robots as the physical embodiment of its Muse AI agent, aiming to bring its intelligence into the real world.

While Meta has no immediate plans to sell these units, it is already taking steps toward industrial scale. The company is hiring in Shanghai and Shenzhen to facilitate mass production of these robots. This move comes as Washington increases scrutiny on the import of new humanoid robots from China, where companies like AgiBot and Unitree currently lead the ecosystem.

Meta's entry into this space is backed by its massive resources, including commitments of $150 billion or more for data centers. By combining its AI model expertise with robotics hardware, Meta is positioning itself as a major player alongside Tesla's Optimus. While mainstream global markets for humanoid robots may be years away, the current focus is on rapid experimentation and infrastructure building.