This year, TIME editors launch the TIME100 Companies: Industry Leaders lists, an expansion of the TIME100 Most Influential Companies issue that dives deeper into 20 sectors to look at the companies shaping their industries. These are the 10 most influential companies in AI of 2026.

ByteDance

From TikTok parent to AI powerhouse

Best known outside China as the company behind TikTok, ByteDance has morphed into an AI-first tech player, with products spanning chatbots, image and video generation, and cloud infrastructure. The foundation of its empire remains Douyin, the Chinese counterpart to TikTok, which boasts roughly 770 million monthly active users. But the clearest sign of where ByteDance is headed is Doubao, its AI assistant, which has amassed more than 155 million weekly active users—and topped 100 million daily during the Lunar New Year holiday in February—making China among the first countries employing AI assistants at mass-market scale. The Beijing-headquartered firm, valued at upwards of $550 billion in recent secondary transactions, reportedly budgeted over $20 billion in capital spending in 2025, most of it on AI infrastructure. ByteDance is also reportedly planning to spend $14 billion on Nvidia chips in 2026, contingent on U.S. export approvals, to maintain its place at the vanguard of China's AI ambitions.  “We feel more and more deeply that AI development is still in its early stages, and at the moment is just the first 500 meters of a marathon,” ByteDance CEO Liang Rubo said in June. —Charlie Campbell

Amazon

Powering the AI arms race

Amazon built its empire on e-commerce and cloud computing. Now it is emerging as one of the central infrastructure players in AI. In late 2025, the company switched on Project Rainier, one of the largest AI computing clusters ever constructed, a vast 1,200-acre complex of sites named after the mountain that looms over Amazon’s Seattle headquarters. The data centers run nearly 500,000 of Amazon's custom Trainium2 chips and help power Anthropic's AI models. Amazon has committed as much as $13 billion to the startup since 2023, with another $20 billion on the table. "We think very highly of the Anthropic team," says Ron Diamant, chief architect of Amazon's Trainium chips. "We give advice on the models that they're building, and they get an early peek into our next generation hardware." Then, in February, Amazon deepened its infrastructure push by pledging up to $50 billion to Anthropic's chief rival OpenAI—conditional on the company using two gigawatts-worth of Trainium infrastructure. "Getting feedback from both is a net positive," says Diamant. Amazon is hedging its bets on model makers as it seeks to control the computing layer AI startups depend on. —Nikita Ostrovsky

Zhipu

No Western chips required

If DeepSeek proved that Chinese firms don't need billions in outside investment to build frontier AI, Zhipu AI is proving they don't need Western chips, either. Founded in 2019 by researchers from Tsinghua University, Zhipu became the first Chinese large language model company to go public, debuting on the Hong Kong Stock Exchange this January after a $558 million IPO. By February it unveiled GLM-5, a 744 billion-parameter model, and GLM-Image, trained entirely on Huawei processors. Released under an open-source license, GLM-5 has surpassed rivals like Google's Gemini 3 Pro on some benchmarks and approaches Claude Opus 4.5 and OpenAI's GPT-5.2 on coding and agentic tasks. Zhipu's models already power more than 4 million enterprise users and developers from 218 countries and regions, and in March 2026, the company announced an annual revenue of 724 million yuan (around $107 million), up 132% from last year. It's one of the strongest signals yet that Chinese AI developers can compete at the frontier without access to Nvidia hardware—or perhaps any reliance on Western technology. —Gabriela Riccardi

OpenAI

Setting the pace

Even by Silicon Valley standards, OpenAI’s scale is staggering. In March, the company closed a $122 billion funding round at an $852 billion valuation. ChatGPT now has more than 900 million weekly active users, and the company’s revenue has climbed to $2 billion a month. Rivals including Anthropic and Google are pressing hard, but OpenAI remains the company setting the pace of AI deployment, with a reach that spans consumers, developers, and businesses. Meanwhile, the company has roiled controversy, signing a Pentagon deal the same day rival Anthropic was banned from government work, and fighting allegations that ChatGPT may have played a role in a suicide and a mass shooting.

To sustain its scale, OpenAI has been building out a sprawling infrastructure network with partners including Microsoft, Amazon, Oracle, SoftBank, and Nvidia, while also trying to tighten its product strategy. In recent weeks, it shut down Sora, its video-generation app, and indefinitely paused plans for an erotic mode as it redirected attention toward products with clearer commercial payoff, especially coding, workplace tools, and enterprise services. OpenAI’s pitch is that ChatGPT’s consumer reach now gives it a direct path into offices and organizations around the world. “We cannot miss this moment because we are distracted by side quests," CEO of AGI deployment Fidji Simo reportedly told employees at an internal meeting in March.  —Tharin Pillay
Disclosure: Investors in OpenAI include Salesforce, where TIME co-chair and owner Marc Benioff is CEO. OpenAI and TIME also have a licensing and technology agreement that allows OpenAI to access TIME’s archives.

Alphabet

The clairvoyant

While Sundar Pichai isn’t a household name like his fellow tech CEOs, he has played a central role in Google’s AI comeback. A decade ago, he declared Google parent company Alphabet would be an “AI-first” company, and began building the foundation that would make that happen. Then, after the rise of rival AI company OpenAI in 2022, he led an internal technological sprint, driven by the company's AI wing Google DeepMind, that catapulted Google’s Gemini towards the top of capability leaderboards. The bet has paid off: Google just crossed $400 billion in annual revenue for the first time, and is the second-most valuable company in the world (depending on the day). No company delivers AI to so many people in so many places, including Gmail, Maps, YouTube, and Waymo. —Andrew R. Chow

Meta

Social media bellwether

The AI boom is a profitable time to be sitting on user data. In 2025, Meta hit record ad revenue, in part by throwing more computing power at the copious data that it has collected through its social networks. And the company is hoping that it can turn the billions of dollars that it’s spent on researchers and new data centers into even more advanced AI models that will “understand people's unique personal goals,” according to Mark Zuckerberg. “We think that the current systems are primitive compared to what will be possible soon,” he said in January.

Perhaps the company already understands its users a little too well. In March, a U.S. jury found Meta (and co-defendant Google) negligent for designing features, including algorithmic recommendation engines, that harmed a young user’s mental health, and awarded the plaintiff $6 million in damages. (Meta said that it “respectfully” disagreed with the verdict and would appeal.) It’s a pittance for the social media behemoth, but with thousands of similar lawsuits pending against Meta and other social media companies, the verdict could set a costly precedent. Still, the company is forging ahead, maintaining its reputation as bellwether for the industry. —Nikita Ostrovsky

Anthropic

Steadfast AI lab

When Anthropic CEO Dario Amodei walked into the Pentagon on February 24, his company's AI model, Claude, was already deep inside the national-security establishment. Anthropic had pushed to get it there, believing that keeping the best AI out of the military's hands would only cede the advantage to America's rivals. The first frontier AI system cleared for classified networks, Claude reportedly had been deployed in the capture of Venezuelan President Nicolás Maduro in January and would be used in the U.S. strikes on Iran days later. But Defense Secretary Pete Hegseth wanted more: he pressed Amodei to drop “red lines” barring Claude's use for mass domestic surveillance and fully autonomous weapons systems, insisting the military should be able to use the model for "all lawful purposes." When Amodei refused, the Trump Administration designated Anthropic a supply-chain risk—an unprecedented move against an American company—and ordered all federal agencies to stop using Claude. Anthropic challenged the designation in court, and a federal judge granted a preliminary injunction blocking it. When the dust settled, Anthropic appeared to have emerged stronger, with a surge of new customers drawn to the company's willingness to hold its ground. That Anthropic had voluntarily put Claude inside the Pentagon in the first place seemed beside the point. —Billy Perrigo
Disclosure: Investors in Anthropic include Salesforce, where TIME co-chair and owner Marc Benioff is CEO.

Alibaba Group

Reshaping more than retail

In less than three years, Chinese tech giant Alibaba has become a dominant force in open-source AI. Its Qwen series has surpassed 1 billion cumulative downloads and spawned more than 200,000 derivative models, making it the world’s most popular open-source model family. Its appeal reaches well beyond China. Airbnb has said it relies heavily on Qwen for its AI customer-service agent, citing the quality and low cost, and Pinterest uses Qwen to analyze visual content and generate contextual text for pins. Alibaba is now trying to turn its open-model lead into a full-stack AI empire. The company is scaling its already expansive cloud computing infrastructure, making its own AI chips and consumer applications, and selling agentic, hosted versions of its models to enterprise clients. "Over the next five years, our goal is to surpass $100 billion in combined cloud and AI external revenue," CEO Eddie Wu told analysts on a recent earnings call. —Tharin Pillay

Mistral

Europe’s AI alternative

As many leading AI companies have retreated from open-source AI—models anyone can download, modify, and run—Mistral has made openness a business strategy. The Paris-based AI company, founded just three years ago and now valued at nearly $14 billion, builds models for coding, transcription, document recognition, and multimodal tasks, and helps customers run them on their own infrastructure, an appealing option for companies and governments that don’t want sensitive data leaving the building. "What matters is for customers to take AI systems and make them their own,” says CEO Arthur Mensch. The strategy is paying off. Mistral’s annualized revenue hit $400 million in early 2026, up roughly twentyfold from the prior year, driven by more than 100 major enterprise clients—including ASML, TotalEnergies, and HSBC—and several European governments drawn to a homegrown alternative to American AI providers. In January, France’s armed forces ministry signed a framework agreement to deploy Mistral on national infrastructure, one of Europe’s first major sovereign AI procurements and a sign that governments increasingly view homegrown AI as a matter of national security. Backed by a €1.7 billion funding round led by Dutch chip-equipment maker ASML, Mistral is now expanding its own computing footprint, with a data-center project planned in Sweden. "Open source is going to win long term,” says Mensch. “It's the one thing that makes economical sense." —Nikita Ostrovsky
Disclosure: Investors in Mistral include Salesforce, where TIME co-chair and owner Marc Benioff is CEO.

Hugging Face

Democratizing AI

With governments and big tech racing to lock up AI capability, Hugging Face is betting the counterweight is openness at scale. The New York-based company operates the largest public repository for AI models, datasets, and applications—more than 2 million models and 500,000 datasets at last count—making it something like the GitHub of AI. More than 30% of Fortune 500 companies have accounts on the platform. “We’re staying on our mission because we think it’s more important than ever,” says CEO Clément Delangue. Hugging Face has also pushed aggressively into AI agents and robotics: its smolagents framework lets developers build tool-using assistants with minimal code, and last year it launched an open-source robot called Reachy Mini. Delangue says agentic AI could reshape Hugging Face, as software agents increasingly fetch models, datasets, and tools on their own, making the platform a resource for both AI builders and the systems they build. “By the end of the year, we expect that we're going to have more agent users than human users,” he says. —Chris Stokel-Walker
Disclosure: Investors in Hugging Face include Salesforce, where TIME co-chair and owner Marc Benioff is CEO.

Correction, April 29:
The original version of this story misstated the model that was trained on domestic chips. It was the GLM-Image model, not the GLM-5 model.