Scale AI

These companies are built to exhibit resilience amid volatility—and deliver better-than-average growth in the long term.

The Future 50

These 50 companies exemplify the innovation, progress, and potential of AI.

AI Innovators 2025-Scale

Fortune 50 AI Innovators

The essential guide to a sector that’s now attracting one out of every four startup investment dollars.

AI-Innovators-2023-Scale

AI automates, but to get to the sophistication of chatbots like ChatGPT requires humans—and a lot of them. While the LLMs behind OpenAI's ChatGPT or Anthropic's Claude learn the statistical connections in language on their own, to guide and refine the responses these models will provide through a chatbot interface, companies turn to a process known as reinforcement learning from human feedback (or RLHF) where human evaluators judge whether the model's answers are useful, helpful, and "safe" (usually meaning not offensive or likely to be used for harm). Then the model is rewarded for producing answers more like those that the human evaluators labelled as good. But all that requires humans. Which is where Scale comes in. Its whole business is around providing the data labels that AI companies need. Often, those labelers work for low wages in the developing world. In Kenya, for example, Scale AI, a Silicon Valley darling, paid workers between $1 and $3 to train AI models, according to The Verge. (A spokesperson said the company adheres to minimum wage law in all counties in which it operates.) Founded by Alexandr Wang in 2016, who was then only 19 years old, Scale is best known as the data maven behind many of the world’s most sophisticated AI models, including those used by a number of self-driving car companies and to support decision-making in defense. The tasks are often simple and monotonous: identify characteristics of retail products or identify objects in an image. While Scale got its start labeling objects in image data for self-driving algorithms, it has now expanded its training to a wide assortment of AI models for customers like Etsy, Instacart, Meta, and, of course, OpenAI. And it’s ridden the AI boom to raise a total of approximately $600 million at a most recent valuation of $7.3 billion. CTO Vijay Karunamurthy said, however, that Scale, which also builds its own AI algorithms, is looking to go beyond finding laypeople to more curated experts, like an authority in ancient Aramaic, to help train models: “Expert feedback is incredibly valuable to help the model structure its thinking.”

Future 50

Overall Score

Sector

Get insights into Scale AI including business profile, stock information, news, and rankings from Fortune.

Scale AI Company Profile, Stock Price, News, Rankings

legacy

scale ai

Companies

Scale AI | Fortune

Scale AI Company Profile, Stock Price, News, Rankings | Fortune

DBS, Southeast Asia’s largest and most profitable bank, is betting big on artificial intelligence to define the next era of global banking. From algorithmic credit scoring and AI-enabled fraud detection to hyper-personalized customer engagement, the bank is embedding intelligence across its platforms to drive both efficiency and growth. With regional competition heating up and digital expectations rising, the challenge is no longer whether to adopt AI—but how to do so responsibly, securely, and at scale. CEO Tan Su Shan shares how the bank is navigating that challenge, what being AI-fueled means in practice, and how institutions across Asia can turn automation and analytics into real strategic advantage.

The Intelligence Dividend: How DBS Is Deploying AI To Reshape Regional Finance

Conferences

The future of autonomous vehicles is here. From Qualcomm’s cutting-edge chips powering next-gen perception to WeRide’s real-world deployments across urban Asia, AI is transforming not just how we drive, but how mobility is imagined, regulated, and scaled. We unpack the breakthroughs and bottlenecks in hardware, data, safety, and trust that will define whether robotaxis become everyday reality or remain stuck in the slow lane.

The Reality Of Robotaxis

Autonomous vehicles have promised a future of convenience, but trust, policy, and performance are still major factors in how mobility is regulated and scaled. Here to discuss the reality of RoboTaxis, please welcome We Ride General Manager of Singapore, Dr. Kerry Xu, Qualcomm Technologies vice president and Q Qualcomm president of Taiwan, Southeast Asia, Australia, and New Zealand, ST Liu, and Fortune Andrew Neska. How fun was that session, right? Um, OK, we're here to talk about autonomous vehicles, robotaxis, if I may, um, so let's level set as I often like to do, um, Carrie, I'll start with you. Where are robotaxis specifically in the region? Where are we seeing them in city streets and where have they not reached yet? OK, um. Uh, I think, uh, a lot of people are already familiar with, uh, like what's happening in the US. So I focused on Asia and just, um, uh, actually I think China is uh really seeing a large scale the robo taxi deployment, uh, in Beijing, Guangzhou, uh, like we ride with in this year at February, we just launched our robo taxi fire charging commercial operations in Beijing. And we now have like more than 50 of the robotaxis providing service and in Guangzhou, same thing, we already launched our commercial charging robo taxi service and most of you may also know Baidu, I think they launched the robotaxi in Wuhan with like 100 of the vehicles. And the second, uh, hot field, uh, in Asia, I think it's Middle East, um, a lot of the you maybe all the red news we are teaming up with Uber to launch our robotaxi actually in Abu Dhabi, Dubai, and in fact we already started the driverless robo taxi test in UAE right now. Wow, that's quite, I mean just thinking about all those cities, they are dramatically different. Urban landscapes and I can't even imagine the technological challenges, but that's precisely what we're going to talk about today. Um, ST, the tech is advancing rapidly. And yet there is so much we can't do. So, again, tell us what's possible, tell us what's on the cusp of changing as it pertains to autonomy. Sure. Uh, so our company comes from uh Years of innovations and creativity in coming up with a computing efficient, very efficient computing. Uh, devices and solutions and from there also with very efficient power consumption, uh, intelligence on the edge and all that. And as the automotive progresses, the need for intelligent processing, very, very high power processing at a very efficient power consumption has become more and more important. Now using that legacy of knowledge and experience over the 40. Years or so, um, we have put together a solution, what we call like Snapdragon chassis, which, which is a solution combining all the different platforms that you would need to make a AD you know automotive driving car, autonomous driving car or AAS or anything that you need to make driving or not to drive such car very comfortable infotainment, telematics and all that. Our role is to provide such a solution so that our customers can put them into products relatively faster and then as the need for autonomous driving cars become more and more important, especially in Southeast Asia or Asia where the driving condition, the weather condition is very, very different, that solution is important because, you know, ouan will bring. And encompassing that solutions, these knowledge and then on top of that allow our OEM to expand onto it, you, you know, especially with AI become available and all that, yeah, so obviously there's a lot of technology that goes into in an autonomous vehicle, right there's the power train, whether it's electric or otherwise they're almost uniformly electric, um, there's. The sensors themselves, which are relatively straightforward affair. One thing I've noticed is, as you said, you know, the kind of ever evolving intelligence of it all perfect for this conference here, you know, it has been clear in Qualcomm's earnings reports, for example, that the company is very bullish on. Artificial intelligence moving to the edge, right, moving to these out of the data centers that we love to talk about at AI conferences like this and into these other places like cars and whatever where the power, as you said, the power and the efficiency and the size, it's all a different consideration set. How is that changing the autonomous market? Well, A lot of this decision making, the decision making process and the speed of this decision making needs to happen locally. You know, latency is extremely important. So intelligence at the at the edge is a very pivotal feature. Uh it's, it's, you know, you, you imagine driving down the road, you know, you have to make decisions very, very quickly instead of going out to the car and come back in again. That's a very critical and a very typical example of intelligence at the edge. That's that's where our forte is, you know, AI at the edge, and that is, uh, that is bringing a lot of benefits from multiple perspectives, latencies one. Um, it also helped to bolster, uh, what do you call it, the protection of information data and so on and so forth. And so that's how it is changing and, and, and with AI coming on board, the, the process of learning, the, the deductions, decision making is becoming more automated and it allows a lot of developers to come up with applications that you can actually use inside the car. So autonomous driving car is not just a car that by itself, but there are lots of things you can do inside the car which bolster the drivers or the passengers' confidence sitting in the car and all that. Yeah, and to do some of that compute um locally is candidly necessary. I live in Los Angeles. There is a mountain range that cuts through the center of LA, uh, in many of the, in many places in that mountain range, there is no cellular service. I could not imagine the car just deciding, ah, we're gonna just give up, you know, mid-drive, um. I want to come back to you, Carrie, um, what are you learning as your company deploys in these different markets? As I said earlier, very different urban landscapes, very different challenges, we mentioned weather earlier, there's the landscape itself, um. There's obviously, uh, uh, uh, you know, kind of, uh, the expectations of the drivers in a given market as well. What are you learning as your company expands? Yeah, so so far we have been deployed our erasing more than 30 cities across 10 countries. We do learn a lot of uh lessons. But I want to emphasize the three lessons we learned. The first is that we, I think we have to work closely with the government and the regulators. We monitor the regulations to make sure we actually conform to the local traffic laws and the regulations. And uh in this part we actually uh hosted many delegations from different countries um like in the a few weeks ago, I think Singapore's acting Minister of Transport Jeffy, he actually led a delegate visited our HQ in Guangzhou and At our at there, he actually announced in the next 5 years, Singapore is going to see many, many AVs in the uh on the streets and they are actually planning to integrate the Aways into the uh public transport system. So I think the government support. You know, working closely with the government is one of the things when you deploy in a different country, when no matter which country you go, the first thing is that you should work closely with the government on regulation. It's an age old technology question, but we've articulated all of the technological, technical challenges of of autonomous driving, but Given what you've just said, in your opinion, do you think the greater hurdle to full deployment are the humans or the tech? Um, from my point of view, I think technology is actually mature enough. Uh, we have seen the like operations of way more in San Francisco, Los Angeles, and like uh we ride, we are operating driverless, no safety operator on board in Beijing, Guangzhou, you know, Abu Dhabi, Dubai, so technology wise. I think it's clearly mature enough to scale, but we have to work with the government on the regulation part, public acceptance, data transparency, all this. I think it takes time to for the community to completely accept the AV as like a part of their normal life. I think it will be. There, but it's just step by step taking some time to achieve it. Everybody remembers their first autonomous ride once you do it, you know, um, SD, I'm gonna, uh, come back to you. You work with many different companies and many different geographies. Uh, same question. Do you feel like the technology is what's, uh, preventing full deployment of these things, or is it in fact the humans, the policy, the. You know, I'm gonna answer this way, we always advocate that um uh that we, we always should have a cross-indust industry benchmark transparency in the training data and uh make sure that we are compliant. To safety, to, uh, to policies, etc. etc. I think those are the things that are built into our architecture. And by doing so, uh, there's you, you build trust in people, you have to build trust in passengers, in policymakers, in general public, etc. etc. So that's part of the importance of building that into the ecosystem and it's not just a technology thing, it is a private sector having to work with the government and uh you know, work with the cities and education and so and so forth, and it comes from both the Tech technological solution for us, it comes in the form of the chip set, the SOCs and the and the software stack that our OEM partners can, you know, choose and, and, and, and, and build into their cars. But at the same time, the build the confidence, I earlier mentioned the things that you can, you can do in the car, significant increase. So take care of the safety, make sure that it is transparent and then you can enjoy yourself while you are in the car, can build up that kind of Acceptability to this whole thing, but I, you know, safety is extremely important, so we have to benchmark, we have to be transparent and so on and so forth. Yeah, yeah, and I'm hearing the word trust quite a bit at this event, so I, I understand. I have one more question for you, but I just want to tell the audience, hey, we're gonna come to you for questions next, so, uh, get ready, um, all right, so. I see you, your organization works with virtually every automaker everywhere and you're of course, you know, ultimately an American company, a global company, um, and you're seeing uh massive differences across those markets in some regard. um I'm curious how you're, you know, you're supplying these, uh, automakers with this technology and working with them to integrate it in the way that it happens. Technology has been incredibly disruptive to the automotive industry. Just speaking from the US perspective, um, Detroit has built a car a certain way for many, many, many years, and it was very, very good at that. And then Silicon Valley shattered it, just totally reversed it, you know what I'm talking about, just completely just rethought the way you would put together, uh, honestly an end point is the way you might think of it in technology, right? ST as this stuff sweeps through, are you finding with your automaker partners a difference in the way they build their vehicles, the way they incorporate this technology, and by the way, is that different in, say, China or Southeast Asia versus the US? I think the big shift to the automotive is one term that you have heard a lot about is like the software defined car. I think universally I can, I can say that all the automakers. Uh, coming to the realization that this is one area that uh everybody needs to get master in. In the past, you look at power trend, you look at the car design, you look at, you know, the things that uh automotive auto automakers will look at. Now your conversation and your considerations uh Kind of get changed a little bit and say, what is your experience when you get into the car and that's where a lot of the differentiations will be brought up as well. And so if I go into the car and then I can conduct my meetings, I can use AI to ask where I am, book my restaurants and all that, just like an extension of my office or my home. I think they will make wonders and do wonders for a particular OEM or any OEM that will, you know, that's one of the ways now to differentiate the vehicles and then on top of that, no one market, no one country, no one geography is the same. So we must enable the continuous learning and with AI being available then now you do the continuous. Learning, use the smart way to factor all these things into the solutions. I think that is the way I would answer these questions. I think the opportunity is there for everyone and I can see very clearly the realization that a software defined car is is is is that is very much with us today, so I just want to correct one thing that you said. You're not going to be going into a self-driving car and booking your meetings. We have AI agents for that. Yeah, yeah, yeah, yeah, we have AI AI that's right. We solve that right here at this event. OK, let's go to the audience for questions about autonomous vehicles and robo taxis taxis. Yes, we have one right here. We just wait for the mic to reach you, and when it does, your name and who you're with. Thank you. Thanks. Hi, uh, my name is Firoz. I'm with Syngenta. Uh, my question is, how important is the infrastructure outside the car? Quality of roads, lane markings, curb, behavior of other drivers on the road, uh, for, for it to be safe. Gentlemen, which one do you want to take that, OK? Uh, yeah, I think that's a very good question. So I think earlier Andrew mentioned is the vehicle itself must be intelligent because, uh, we put the like brains in the vehicle, but Infrastructure is also important. For example, uh, network connection is like 5G, uh, a lot of this, uh, autonomous vehicles right now still have like remote monitoring in case the vehicle stuck or in certain. Conditions, the people who are monitoring remotely can still give a hint or help the vehicle. Also, in Singapore, we are already working with the government on like how to design the pickup job point to be friendly to Aways. And how you uh monitor all the different uh pickup points to uh be able to uh see where where the vehicles, all these kinds of uh infrastructure, even the how to educate the people to better interact with the AVs. I think this is something a lot of government. Actually already start looking at and I have worked with some scholars whose study topic is how to make our infrastructure friendly to uh driverless vehicles. I think this is something is going to be happening in the next uh maybe a few years. Yeah. Very good. All right. Another question. Yes, we have one over here, right here. Just wait a moment again, your name and who you're with. My name is Selvin. I'm with Alliance Asia. Just have a question. Do you think that LIDA technologies is necessary to make autonomous vehicles more safe? Is light are necessary to make it. Yeah, um, I think, uh, from my point of view, I think, uh, LEA is very necessary. Uh, in fact, is, uh, my, my, my idea is the reason some companies don't install the ladder is probably some a few years ago, ladder is still pretty expensive, uh, sensors to install which increases the cost for the vehicles. However, if you look at the edge case, so the vehicles are different from like uh some other like your cellphone. Uh, tolerance is different for vehicles. We are not looking at 99% safety. We're probably looking at 99.9999 safety. We don't, we can't afford the accident due to age case. You cannot catch it. So cameras has uh vision based has some limitations like in the very low light environment like midnight, no light, but lighter. You know, it's uh it's emitted the light, so it doesn't rely on any lights and also the lighter can accurately detect the shape of the obstacles which also helps the algorithm to more advance the control and planning algorithm to bypass or to avoid the obstacles. So with the lighter price is cheaper and cheaper, I think. Uh, like us, we are doing is we do the fusion between the camera and the ladder. So all this allows us to detect all the obstacles in like 200 m, 300 degrees, very accurately to do the planning and controlling. So yeah, that's how I believe the ladar is still very necessary. If I just add to the conversation about the infrastructure around the vehicles. I think one of the key things going into the future is. So the cellular V to X, cellular equal to any everything. And then if you think about the future of autonomous driving, uh, CV to X is something that we in Qualcomm, uh, continue to, uh, to invest a lot in. And to your point of light, when you have CV to X, obviously you can plan ahead on many, many things and and so and so forth. So I think, uh, to the questions about the road marks and things, CV. To ask the road maps and all these, just point to the fact that we need to work very closely with infrastructure partners as well as ecosystem partners and all that. Yeah, in terms of the infrastructure, I have a very good example. I keep telling my friends is that right now in the car park in certain areas we install mirror for the help drivers to look at the blind spot, but in the future for avis actually we actually can install. Those B2X equipment to help the AVs to look at the black spots. I think these are all the future probably development for the infrastructure to be AV friendly. But it is unquestionable that our our urban landscapes are going to change dramatically because of of the rise of robotaxis and autonomous vehicles. So gentlemen, thank you so much for an enlightening conversation, a round of applause please for our panelists. Thank you. Thank you.

As artificial intelligence reshapes the foundations of innovation, how we design, fund, and govern technology has never mattered more. We’ll talk with a policy strategist, a systems thinker, and a market-focused investor to explore how our collective choices can shape AI not just for scale—but for human flourishing. From institutional trust and design frameworks to capital flows and societal risk, we'll unpack what it means to invest in both human and machine intelligence in ways that are ethical, adaptive, and imaginative. At the crux of this conversation is creativity itself: how can generative AI unlock new forms of expression, co-creation, and cultural value—and what it will take to fund and govern those possibilities responsibly?

Intelligence by Design: Aligning Policy, Capital, and Creativity in the Age of AI

As climate risk, supply chain disruption, and population growth strain global food systems, the race is on to cultivate smarter, more resilient agriculture. From AI-powered greenhouses and robotic harvesting to digital twin farms and gene-edited seeds, this session explores how technology is transforming every layer of the agri-food chain. We’ll hear from leaders at the cutting edge of sustainable agriculture, data-driven farming, and food innovation as they unpack what it takes to feed the future—ethically, efficiently and at scale.

Scale AI

Scale AI in Fortune Rankings

Videos