At almost the same moment Nvidia CEO Jensen Huang was reacting to OpenAI’s surprising partnership with AMD, Sam Altman was already teasing more major deals ahead.
Huang, appearing on CNBC’s Squawk Box, admitted he had little knowledge of the AMD announcement, which gives OpenAI the chance to acquire up to 10 percent of AMD’s stock in exchange for helping develop its next-generation AI chips. The deal effectively makes OpenAI a shareholder in the chipmaker.
Nvidia’s arrangement with OpenAI takes the opposite form. After agreeing to invest as much as $100 billion into the AI giant, Nvidia has become one of OpenAI’s biggest backers, supplying it directly with advanced hardware for the first time.
Huang said the new sales approach will allow OpenAI to build its own data centers, describing it as a step toward becoming a “self-hosted hyperscaler.” But he admitted the cost is massive—each gigawatt of AI infrastructure could run between $50 billion and $60 billion.
In 2025 alone, OpenAI has commissioned 10 gigawatts of U.S. facilities through its $500 billion Stargate project with Oracle and SoftBank, alongside a $300 billion cloud deal with Oracle. Its partnerships with Nvidia and AMD add another 16 gigawatts of capacity, while expansion continues in the U.K. and across Europe.
During an interview on Andreessen Horowitz’s a16z Podcast, Altman said these deals are only the beginning. Confident in OpenAI’s upcoming models and technology roadmap, he hinted that the company will pursue even more collaborations in the coming months.
“To make the bet at this scale, we need the whole industry to support it,” Altman said, noting that OpenAI’s infrastructure race will rely on deep partnerships “from electrons to model distribution.”
OpenAI’s billion-dollar momentum, it seems, is only just getting started.
OpenAI, the company behind ChatGPT, is making big moves in India. Here’s what you need to know about its plans for the $500 billion Stargate supercomputing project and how it could reshape India’s data center landscape.
1. OpenAI Is Talking to Major Indian Data Center Players OpenAI is in discussions with top Indian firms including Sify Technologies, Yotta Data Services, E2E Networks, and CtrlS Datacenters. These talks are part of OpenAI’s strategy to bring its Stargate project—a $500 billion global AI infrastructure initiative—to India.
2. Reliance Industries Plans the World’s Largest Data Center Reliance Industries, known for its oil, telecom, and digital businesses, is building a massive data center in Jamnagar, Gujarat. The facility could align with OpenAI’s Stargate requirements, making it a potential partner in the project.
3. Key Considerations for OpenAI Partnerships The discussions focus on critical factors such as installed capacities, geographical spread, and reliable power availability. These elements are essential for handling the high-performance AI workloads that Stargate will require.
4. India Wants Data Stored Locally The Indian government has requested OpenAI to invest billions locally and store Indian data within the country. This push aligns with India’s broader strategy to localize data storage and processing, similar to initiatives by Microsoft and Google.
5. OpenAI Is Expanding Its Presence in India CEO Sam Altman recently noted that India is OpenAI’s second-largest market after the US, with potential to become the largest. The company is opening its first office in New Delhi and has enabled local data residency for AI interactions, ensuring that prompts, uploads, and chats stay in-country.
6. Stargate Project Explained Launched in January, Stargate is a $500 billion joint venture investing in AI infrastructure over four years. Initial backers include SoftBank, Oracle, MGX, and OpenAI itself. Technology partners such as Arm, Microsoft, NVIDIA, and Oracle are central to building the supercomputing backbone.
7. Local Infrastructure Challenges and Opportunities Experts say Indian data centers have the talent, power, and market demand to meet AI workload needs. However, specialized AI hardware and advanced cooling solutions remain supply chain challenges. Strategic partnerships and targeted investments could help bridge the gap and make India a core hub for AI infrastructure.
OpenAI’s potential entry into India signals a major step in the country’s AI journey, promising cutting-edge infrastructure, local job creation, and a bigger role in the global AI landscape.
Thursday, August 7, 2025 – Launch Day OpenAI unveils GPT-5 during a live presentation, showcasing its new “real-time router” that decides whether to respond quickly or spend more time thinking through answers.
During the event, a bar chart appears showing a lower benchmark score with a much taller bar, sparking immediate confusion and jokes online. The term “chart crime” begins trending among AI watchers.
CEO Sam Altman later posts on X, calling it a “mega chart screwup” and clarifying that the correct charts were in the official blog post.
Thursday Afternoon Early users begin noticing GPT-5 responses feel less capable than GPT-4o. The culprit: a server issue disables the autoswitcher, meaning the model’s intended intelligence boost isn’t kicking in.
Friday Morning – Reddit AMA Begins Sam Altman and members of the GPT-5 team host a Reddit ask-me-anything session on r/ChatGPT. Users flood the thread with questions about the rocky rollout, router problems, and the infamous chart.
Friday, Midday Altman explains the technical glitch, confirming the router problem was fixed and promising tweaks to make the model-selection process more accurate. He pledges more transparency about which model is responding to each prompt.
Friday, Afternoon Responding to repeated user requests, Altman says the team is “looking into” letting Plus subscribers continue to use GPT-4o. He also promises to double rate limits for Plus users as rollout completes.
Friday Evening – Closing Words Altman ends the AMA with a pledge to “get things stable” and keep listening to feedback. Early reviewers like Simon Willison note GPT-5 is strong overall but still struggles with tasks like accurately formatting tables.
Aftermath The “chart crime” becomes a meme in the tech community, with jokes about letting GPT itself make PowerPoint slides. For now, GPT-5 continues to evolve as OpenAI works to regain user confidence after its messy debut.
1. ChatGPT is on pace to reach 700 million weekly users this week. OpenAI’s consumer-facing chatbot continues its explosive growth, with VP Nick Turley confirming the milestone in a post on X.
2. Usage has quadrupled in the past year. According to Turley, ChatGPT has grown 4x year-over-year, reflecting a steep rise in global adoption.
3. The app previously reached 500 million weekly users in March. Just five months ago, ChatGPT hit the half-billion mark, with usage accelerating significantly since.
4. New features have fueled a fresh wave of engagement. In March, OpenAI rolled out upgraded image generation powered by GPT-4, leading to a surge in user interaction.
5. Over 700 million AI images were created within days. OpenAI COO Brad Lightcap reported in April that more than 130 million users generated hundreds of millions of images shortly after the feature’s release.
6. ChatGPT’s paying business user base has jumped. Lightcap revealed last week that the app now has 5 million enterprise users, up from 3 million in June.
7. The app’s stickiness rivals that of major tech platforms. Sensor Tower reported that users accessed ChatGPT more than 12 days per month on average, trailing only Google and X.
8. Users spend an average of 16 minutes daily on the app. In the first half of 2025, ChatGPT users clocked nearly 16 minutes per day, indicating deepening engagement.
9. ChatGPT has rapidly matured into a core productivity tool. The app has transitioned from novelty to necessity for many, becoming integral in education, research, design, and workplace productivity.
10. OpenAI says it remains focused on accessibility and impact. Turley emphasized the company’s goal of making AI useful and widely available, crediting the team for delivering on that mission.
As AI continues to permeate everyday life, ChatGPT’s meteoric growth suggests it’s no longer just a tech trend but a platform shaping how people learn, work, and create.
In a twist so shocking it almost feels predictable, OpenAI CEO Sam Altman announced Friday that the company is once again delaying the release of its much-hyped open model — this time indefinitely — citing the need for more “safety testing.”
The model was originally scheduled to drop earlier this summer, then postponed to next week, and now it’s been yeeted into the future with no ETA in sight.
“We need time to run additional safety tests and review high-risk areas. We are not yet sure how long it will take us,” Altman posted on X, formerly Twitter, as if to reassure developers that everything is fine, just fine.
“While we trust the community will build great things with this model, once weights are out, they can’t be pulled back,” he added, in what might be the understatement of the year.
The open model is supposed to be OpenAI’s grand re-entry into the free-range AI ecosystem — a rare gesture from a company that has spent the last few years locking its best tech behind increasingly monetized APIs and partner deals.
Unlike the upcoming GPT-5, which is still wrapped in secrecy and likely to be subscription-only, the open model was expected to be downloadable and runnable on local machines.
But now, that promise is trapped in limbo.
OpenAI insists the delay is all in the name of responsibility — because nothing says “we totally have this under control” like postponing a product multiple times after hyping it as the summer’s blockbuster AI release.
Aidan Clark, OpenAI’s vice president of research leading the open model team, echoed the vague sentiment in a post of his own on Friday.
“Capability-wise, we think the model is phenomenal — but our bar for an open source model is high and we think we need some more time to make sure we’re releasing a model we’re proud of along every axis,” he said, without explaining what exactly those axes are or what OpenAI is even afraid of.
This isn’t the first time OpenAI has played coy with breakthroughs.
Back in June, Altman vaguely teased that the company had achieved something “unexpected and quite amazing” with the model — without saying what that was.
Apparently, “amazing” now means “not ready for public consumption.”
Meanwhile, competitors aren’t slowing down to overthink it.
Earlier on Friday, Chinese AI startup Moonshot AI dropped Kimi K2, a one-trillion-parameter model that not only calls itself open but also reportedly outperforms OpenAI’s GPT-4.1 on several advanced coding benchmarks.
So while OpenAI dithers over what constitutes a “safe” model, others are shipping — and, crucially, releasing — actual products.
But now, with delays stacking and rivals flexing, that title seems increasingly at risk.
There have also been internal discussions about allowing the open model to interface with OpenAI’s cloud models for more complex tasks.
Yet even those speculative features are now in question.
For developers, researchers, and tinkerers eager to get their hands on a truly open model from the company that literally has “Open” in its name, the message is clear: wait longer, wait quietly, and definitely don’t ask too many questions.
Because in the world of frontier AI, where transparency was once the mission, delays and vague statements are the new normal — and “open” is starting to sound like a brand, not a principle.
OpenAI’s ChatGPT—known best for churning out last-minute essays, barely-passable code, and dodgy diet plans—is apparently trying to grow a conscience.
A bizarre new tool labeled “Study Together” has mysteriously appeared in the dropdown menus of some ChatGPT subscribers, prompting immediate confusion, cautious optimism, and mild existential dread.
According to early reports from confused users on social media, “Study Together” is less about spoon-feeding answers and more about interrogating users with questions.
Yes, you read that right.
Rather than doing your work for you, the chatbot now wants you to think.
This is either the future of responsible edtech—or a sick joke on procrastinators.
OpenAI hasn’t officially acknowledged the feature’s existence, because of course it hasn’t.
However, ChatGPT itself said, “OpenAI hasn’t officially announced when or if Study Together will be available to all users — or if it will require ChatGPT Plus,” which is as close to corporate ghosting as you can get from an AI.
The “Study Together” mode has been described as ChatGPT’s version of Google’s LearnLM, a product Google touted in May 2024 as the next evolution in AI-powered learning.
LearnLM takes a more Socratic approach, pushing students to explain concepts, not just consume them.
It appears OpenAI is following suit—perhaps begrudgingly—after its tool spent the last two years being either a digital tutor or a glorified cheat sheet.
At its best, ChatGPT has helped teachers build lesson plans and aided students in understanding complex topics.
At its worst, it’s become the go-to accomplice for crafting AI-generated term papers, solving math problems without showing work, and fueling a cottage industry of academic dishonesty.
The arrival of “Study Together” might be OpenAI’s attempt to steer the ship back toward productive waters—or at least avoid another angry faculty senate resolution.
The name “Study Together” has sparked speculation that the feature might allow multiple users in a single chat, forming actual study groups guided by the AI.
It’s a cute idea, in theory.
But given OpenAI’s recent moves—locking more and more behind its Plus paywall—it’s unclear whether this mode will be a free-for-all or just another velvet rope reserved for paying customers.
Education experts have long debated whether AI tools are a blessing or a nuclear bomb dropped on academic integrity.
“Students are outsourcing their brains,” said Ethan Mollick, a Wharton professor and frequent commentator on AI in education, in a 2023 talk.
A feature that flips the dynamic—forcing users to engage rather than coast—might just be the first sign that OpenAI isn’t thrilled about becoming the villain in every “AI ruined school” think piece.
Or maybe it’s a PR patch slapped onto a platform that made cheating so easy even your grandma could do it.
Critics are already poking holes in the premise.
“Why would I pay for AI to ask me questions I don’t know how to answer?” quipped one Reddit user.
Another mused that the tool sounds suspiciously like “homework with extra steps.”
Still, in an age when generative AI is being integrated into nearly every level of the education system—from personalized learning to test prep—the idea of AI that builds cognitive muscle instead of enabling mental shortcuts is intriguing.
It’s a noble goal, if not one destined for instant popularity among users who just want the answers five minutes before class.
For now, “Study Together” remains in the experimental shadows, like a digital after-school special nobody asked for but someone somewhere thinks we need.
And if it catches on, it might just be the beginning of AI asking us to pull our own intellectual weight—for once.
OpenAI, the darling of the AI boom and apparent new poster child for corporate anxiety, is doubling down on security like a tech bro hoarding crypto during a market crash.
According to a report from the Financial Times, OpenAI has revamped its internal security protocols, reportedly in a panic following the release of a rival model by Chinese startup DeepSeek in January 2025.
OpenAI believes DeepSeek used “distillation” — a technique as controversial as it is technically legal — to copy its work, effectively serving Silicon Valley’s worst nightmare: intellectual property theft with a side of global competition.
In response, OpenAI’s leadership hit the big red button marked “paranoia.”
The company has implemented what sources dub “information tenting,” a charmingly dystopian term that means only a select few employees are allowed to know what’s going on — even inside their own offices.
During the development of the mysterious o1 model, only team members who were explicitly “read in” could even talk about it, and only in approved locations.
Presumably, the watercooler is now classified territory.
If that wasn’t enough, the company has taken to storing some of its most valuable tech in offline computer systems.
That’s right — the AI revolution now includes air-gapped vaults.
Biometric scanners have also been introduced, because nothing says “trust” like making your employees scan their fingerprints just to enter a room.
And if you want to connect to the internet?
Better file a request and wait — because OpenAI has introduced a “deny-by-default” policy for all outbound connections.
The goal here, we’re told, is to fend off foreign adversaries.
But let’s not kid ourselves.
Given the backstabbing talent wars between major U.S. tech firms and the increasingly leaky nature of OpenAI’s own internal communications — including juicy off-the-cuff remarks from CEO Sam Altman — the company might be as worried about insiders as it is about any shadowy overseas operatives.
Security at OpenAI’s data centers has reportedly been ratcheted up too, with extra guards, surveillance, and what one imagines is a growing sense of unease.
Cybersecurity personnel are being hired in droves, perhaps confirming that no one really knows how secure these systems are — or who’s already poking around in them.
The rise of “distillation” as a threat has only amplified these concerns.
The technique, which allows smaller models to be trained on the outputs of larger ones, has been hailed by researchers as efficient — and reviled by companies like OpenAI when it’s used without permission.
In short, the gloves are off in the global AI arms race.
Everyone wants to build the smartest model on Earth, and no one wants to share their notes.
OpenAI, for its part, has remained tight-lipped about the specifics of these security measures, and did not immediately respond to requests for comment.
That’s probably intentional — in this new world order, loose lips sink language models.
Whether all these internal lockdowns will actually stop competitors from replicating OpenAI’s tech is an open question.
But what’s clear is that the company has taken a sharp turn into cloak-and-dagger territory, replacing trust and transparency with clearance levels and fingerprint scans.
Because nothing fosters innovation like making your office feel like a scene from a spy thriller — just with fewer martinis and more keycards.
OpenAI is scrambling to retain talent after Meta reportedly lured away eight senior researchers in a Silicon Valley brain drain standoff.
Chief Research Officer Mark Chen likened the situation to a break-in, expressing outrage in an internal memo leaked to Wired.
In a desperate bid to stop the bleeding, OpenAI executives are reportedly recalibrating compensation and offering “creative” rewards to keep staff from jumping ship.
CEO Sam Altman recently complained about Meta dangling $100 million signing bonuses, a claim Meta has internally denied but which sparked renewed panic at OpenAI.
Leadership says it’s working overtime to plug the talent leak and reassert its dominance in AI before the house gets looted even further.
As artificial intelligence rapidly transforms industries, venture capital firms are at the forefront—fueling innovation, shaping startups, and amplifying the AI race. While the market is now more selective than in the peak hype years, elite investors continue to deploy massive funds into AI infrastructure, applications, and foundational technologies. Here are the 12 venture capital firms currently leading the AI gold rush.
1. Andreessen Horowitz (a16z)
Andreessen Horowitz has become synonymous with the AI boom, thanks to its aggressive pursuit of both infrastructure and consumer-facing AI startups. With broad bets across generative models, developer platforms, and enterprise tools, a16z is positioning itself as the cornerstone of the next wave of AI-native businesses. The firm focuses on long-term disruptive potential and isn’t afraid to go early and big.
2. Sequoia Capital
Sequoia continues to dominate the AI landscape by backing category-defining companies. With investments in both foundational model developers and AI tools built for enterprise efficiency, the firm is adopting a full-stack AI strategy. It’s not just about identifying unicorns—it’s about owning the ecosystem that powers them.
3. SoftBank Vision Fund
SoftBank is back in force, shifting its thesis from broad tech to a tight focus on AI. The firm is funneling its capital into robotics, automation, and synthetic intelligence platforms that aim to scale human output. SoftBank has the capital to lead mega-rounds and the conviction to support bold, untested ideas.
4. Lightspeed Venture Partners
Lightspeed is doubling down on infrastructure-layer investments—AI data platforms, cloud-native tools, and productivity AI. The firm balances early-stage experimentation with scale-up investments, giving it exposure across the AI lifecycle. Lightspeed’s strategy centers on deployability, speed, and enterprise-grade solutions.
5. Khosla Ventures
Khosla Ventures has built a reputation for supporting moonshot AI ideas. The firm invests heavily in deep science, AGI research, AI-in-healthcare, and even AI bioengineering. With a high risk tolerance and long time horizon, it aims to capture outsized returns from the frontier of intelligence.
6. Tiger Global Management
Tiger Global is among the fastest-moving firms in the AI space. While its tech bets span sectors, AI has become one of its highest-conviction themes. The firm pursues companies with explosive growth potential and backs them aggressively across multiple rounds, from seed to pre-IPO.
7. GV (Google Ventures)
GV leverages its deep technical insight to invest in both applied AI startups and cutting-edge research projects. Its internal access to technical talent and domain expertise gives it an edge in evaluating highly technical teams. GV focuses on startups that embed AI into core workflows rather than surface-level integrations.
8. Greylock Partners
Greylock focuses on AI’s impact on enterprise workflows—automation, data intelligence, and hiring tech. The firm supports startups that are rebuilding business infrastructure with AI at the center. It’s especially active in early-stage investments that aim to define new categories.
9. Bessemer Venture Partners
Bessemer brings a disciplined and methodical approach to AI investing. With deep roots in enterprise software, the firm is backing AI tools that improve business productivity, security, and analytics. Bessemer’s AI investments tend to focus on long-term monetization models and product-market fit.
10. Lux Capital
Lux Capital’s AI thesis is centered on deep tech—autonomous systems, computational biology, and defense applications. The firm invests in companies that straddle physical and digital worlds, where AI is applied to solve highly complex, often mission-critical problems. Its contrarian approach gives it exposure to AI’s high-stakes use cases.
11. Conviction (Sarah Guo)
Conviction is a newer firm but already a rising force in AI investing. With a founder-first strategy, it backs early-stage startups building software-native intelligence layers. The firm focuses on infrastructure, agentic software, and developer-first platforms that represent the next wave of AI utility.
12. Intel Capital
Intel Capital has transformed from a corporate VC to an independent powerhouse, backing AI at the hardware, data, and application layers. The firm supports companies innovating in model optimization, edge inference, AI toolkits, and intelligent data systems. Intel’s long-term strategy is tied closely to enabling compute-heavy use cases.
Trends Shaping the AI VC Landscape
The AI gold rush isn’t evenly distributed. A handful of top firms now account for most mega-deals and seed activity. But this new wave isn’t just about scale—it’s about depth, conviction, and clear technological theses.
1. Bigger Bets, Fewer Deals Firms are writing larger checks into fewer startups with stronger technical defensibility. AI founders with deep research backgrounds and differentiated architectures have the upper hand.
2. Infrastructure-First Mindset While consumer AI is exciting, most leading VCs are focused on the picks-and-shovels layer—model optimization, orchestration, data pipelines, and compute efficiency. The assumption is simple: every successful AI company will rely on best-in-class infrastructure.
3. Cross-Sector Explosion AI is no longer confined to tech. Top firms are funding AI in healthcare, finance, logistics, defense, and legal. As the technology becomes horizontal, VCs are assembling portfolios that span both core research and applied industry products.
4. The AGI Horizon Several of these firms—especially those led by technical founders—are actively positioning around AGI (artificial general intelligence). Whether through direct investment or via internal R&D alliances, the long-term bet is that AI will become a new class of infrastructure, like electricity or the internet.
The venture capital firms driving the AI gold rush are not merely financiers—they are architects of the future digital economy. Their capital shapes research labs, enables moonshots, and brings to market technologies that may soon redefine productivity, intelligence, and industry itself.
In this moment, capital is not just chasing returns. It’s picking sides in a global arms race for algorithmic advantage. And for the firms listed here, the stakes—and the rewards—have never been higher.
Legislation introduced Wednesday in Congress aims to ban all Chinese artificial intelligence systems from federal government use, citing national security threats.
The bipartisan proposal, pushed by both House and Senate members, would bar any AI developed by U.S. adversaries, notably China, from operating in federal agencies.
The bill makes rare exceptions for counterterrorism and research purposes, giving the government a free pass to use the tech when convenient.
The legislative panic follows the release of DeepSeek, a Chinese AI model nearly matching OpenAI and Google in power—despite being built on a budget.
Lawmakers say the U.S. must win what they describe as a “new Cold War,” with AI now framed as the world’s most important battleground.
The proposal coincides with increasing concern over China’s growing AI capabilities, as Stanford’s 2025 AI Index shows China catching up in performance and leading in publications and patents.
Lawmakers argue that authoritarian-built AI could “infect” democratic institutions, assuming that technology inherits political values like an ideological virus.
Committee hearings featured calls for stricter export controls on advanced chips, with experts claiming that without limits, China could use U.S. technology to “build powerful AI to harm American interests.”
The bill would empower the Federal Acquisition Security Council to identify AI systems considered high-risk, although no one seems to know how many are already in use.
Federal agencies and some states have reportedly banned DeepSeek from government devices, but the new bill seeks to enforce a sweeping, standardized prohibition.
Despite the sense of urgency, experts note loopholes in existing chip export controls that continue to allow China access to restricted technologies.
Lawmakers insist that AI developed by authoritarian regimes has “no business” operating in American government, even though the technology has likely already slipped through the cracks.
As the AI arms race intensifies, Congress appears committed to securing America’s digital turf—just as soon as they finish figuring out what AI actually is.
