Past companies of the day
The last 60 days, newest first. Each date opens the file that was featured that day.
Anthropic
Founded by researchers who left OpenAI over how it was approaching safety, Anthropic built Claude into a genuine rival to ChatGPT while making AI safety research part of its public commercial pitch rather than a side project.
xAI
xAI's rapid rise and its February 2026 absorption into SpaceX is one of the clearest signals yet that frontier AI labs are consolidating into companies with the balance sheets to fund trillion-dollar compute build-outs.
Together AI
Together AI is one of the clearest bets that open-weight models need their own specialized infrastructure layer rather than running on general-purpose cloud, and its Aramco-led round shows sovereign capital moving directly into the compute layer of AI rather than only into model companies.
Stability AI
Stable Diffusion's 2022 open-weight release is the reason so much of today's AI image-generation tooling exists at all - and the company's own turbulent finances and 2024 leadership overhaul is a cautionary case study in what open distribution does and doesn't do for a company's own balance sheet.
Scale AI
Nearly every frontier model released since 2020 was trained in part on data Scale AI's workforce labeled or ranked, and Meta paying $14.3 billion for a minority stake - and hiring its founder outright - was one of the clearest signals yet of how much Big Tech values control over AI training data.
Runway
Runway pushed generative video from a research curiosity into a tool film and advertising studios actually use, and its shift toward "world models" signals where the company thinks generative media is headed next - simulation, not just clip generation.
Presight AI
Presight is the one piece of G42's empire the public can actually see the books of - a listed, quarterly-reporting company that turns the UAE's sovereign AI ambitions into an audited revenue line, and one of very few AI-native companies of its kind listed on any exchange in the region.
Perplexity AI
Perplexity bet that people would rather read a synthesized, cited answer than click through a page of blue links, and grew from a $121 million valuation to roughly $23 billion in under three years betting on it.
OpenAI
ChatGPT is the product that took generative AI from a research demo to something hundreds of millions of people use weekly, and OpenAI's release cadence has set the industry's pace for most of the years since.
Mistral AI
The only European foundation-model lab operating at a scale comparable to the major US and Chinese labs, and the company most often cited when European governments talk about AI sovereignty that doesn't run through Silicon Valley.
Meta Superintelligence Labs
The clearest example of a Big Tech incumbent restructuring around the foundation-model race in real time - built by pulling in a rival's founder, and quietly abandoning the open-weight strategy that made Llama famous.
Hugging Face
Almost every open-weight model released by a foundation-model lab - Llama, Mistral, DeepSeek included - gets distributed through Hugging Face's Hub, making it the closest thing the AI industry has to shared infrastructure, and Nvidia's 2026 acquisition puts that infrastructure inside the company that supplies most of the industry's chips.
Google DeepMind
The lab behind AlphaFold and Gemini, and the clearest example of how the foundation-model race reorganised even Alphabet's own AI operation - the 2023 Brain/DeepMind merger, and the August 2026 leadership reshuffle under pressure from OpenAI and Anthropic.
G42
G42 is the anchor of the UAE's bet that a small, oil-wealthy nation can become a genuine AI infrastructure power - through direct partnerships with OpenAI, Microsoft, Nvidia and, after 2024, the US government itself - rather than only a customer of AI built elsewhere.
ElevenLabs
A huge share of the synthetic voices you now hear in AI products - audiobooks, dubbed video, voice assistants - were generated by ElevenLabs' models, and its rapid re-rating to an $11 billion valuation reflects how central voice has become to the AI product stack.
DeepSeek
DeepSeek's R1 model forced every major AI lab to publicly defend its compute spending in early 2025, and remains the clearest evidence that frontier-level AI research is no longer a US-only contest.
Databricks
Built on the open-source project that made large-scale data processing practical, Databricks has repositioned itself as the platform enterprises use to actually build AI on top of their own data - a less visible but arguably more durable business than a consumer chatbot.
Core42
Core42 is the physical layer under the UAE's AI ambitions - the data centres, GPUs and sovereign-cloud contracts that make "AI made in the UAE" a real infrastructure claim rather than a slogan, and increasingly a US-based infrastructure operator in its own right.
Cohere
Founded by a co-author of the paper that made modern AI possible, Cohere bet on enterprise and government customers over consumer chat - a narrower, less visible strategy than OpenAI's or Anthropic's, aimed at organisations that need models running on their own infrastructure.
Anthropic
Founded by researchers who left OpenAI over how it was approaching safety, Anthropic built Claude into a genuine rival to ChatGPT while making AI safety research part of its public commercial pitch rather than a side project.
xAI
xAI's rapid rise and its February 2026 absorption into SpaceX is one of the clearest signals yet that frontier AI labs are consolidating into companies with the balance sheets to fund trillion-dollar compute build-outs.
Together AI
Together AI is one of the clearest bets that open-weight models need their own specialized infrastructure layer rather than running on general-purpose cloud, and its Aramco-led round shows sovereign capital moving directly into the compute layer of AI rather than only into model companies.
Stability AI
Stable Diffusion's 2022 open-weight release is the reason so much of today's AI image-generation tooling exists at all - and the company's own turbulent finances and 2024 leadership overhaul is a cautionary case study in what open distribution does and doesn't do for a company's own balance sheet.
Scale AI
Nearly every frontier model released since 2020 was trained in part on data Scale AI's workforce labeled or ranked, and Meta paying $14.3 billion for a minority stake - and hiring its founder outright - was one of the clearest signals yet of how much Big Tech values control over AI training data.
Runway
Runway pushed generative video from a research curiosity into a tool film and advertising studios actually use, and its shift toward "world models" signals where the company thinks generative media is headed next - simulation, not just clip generation.
Presight AI
Presight is the one piece of G42's empire the public can actually see the books of - a listed, quarterly-reporting company that turns the UAE's sovereign AI ambitions into an audited revenue line, and one of very few AI-native companies of its kind listed on any exchange in the region.
Perplexity AI
Perplexity bet that people would rather read a synthesized, cited answer than click through a page of blue links, and grew from a $121 million valuation to roughly $23 billion in under three years betting on it.
OpenAI
ChatGPT is the product that took generative AI from a research demo to something hundreds of millions of people use weekly, and OpenAI's release cadence has set the industry's pace for most of the years since.
Mistral AI
The only European foundation-model lab operating at a scale comparable to the major US and Chinese labs, and the company most often cited when European governments talk about AI sovereignty that doesn't run through Silicon Valley.
Meta Superintelligence Labs
The clearest example of a Big Tech incumbent restructuring around the foundation-model race in real time - built by pulling in a rival's founder, and quietly abandoning the open-weight strategy that made Llama famous.
Hugging Face
Almost every open-weight model released by a foundation-model lab - Llama, Mistral, DeepSeek included - gets distributed through Hugging Face's Hub, making it the closest thing the AI industry has to shared infrastructure, and Nvidia's 2026 acquisition puts that infrastructure inside the company that supplies most of the industry's chips.
Google DeepMind
The lab behind AlphaFold and Gemini, and the clearest example of how the foundation-model race reorganised even Alphabet's own AI operation - the 2023 Brain/DeepMind merger, and the August 2026 leadership reshuffle under pressure from OpenAI and Anthropic.
G42
G42 is the anchor of the UAE's bet that a small, oil-wealthy nation can become a genuine AI infrastructure power - through direct partnerships with OpenAI, Microsoft, Nvidia and, after 2024, the US government itself - rather than only a customer of AI built elsewhere.
ElevenLabs
A huge share of the synthetic voices you now hear in AI products - audiobooks, dubbed video, voice assistants - were generated by ElevenLabs' models, and its rapid re-rating to an $11 billion valuation reflects how central voice has become to the AI product stack.
DeepSeek
DeepSeek's R1 model forced every major AI lab to publicly defend its compute spending in early 2025, and remains the clearest evidence that frontier-level AI research is no longer a US-only contest.
Databricks
Built on the open-source project that made large-scale data processing practical, Databricks has repositioned itself as the platform enterprises use to actually build AI on top of their own data - a less visible but arguably more durable business than a consumer chatbot.
Core42
Core42 is the physical layer under the UAE's AI ambitions - the data centres, GPUs and sovereign-cloud contracts that make "AI made in the UAE" a real infrastructure claim rather than a slogan, and increasingly a US-based infrastructure operator in its own right.
Cohere
Founded by a co-author of the paper that made modern AI possible, Cohere bet on enterprise and government customers over consumer chat - a narrower, less visible strategy than OpenAI's or Anthropic's, aimed at organisations that need models running on their own infrastructure.
Anthropic
Founded by researchers who left OpenAI over how it was approaching safety, Anthropic built Claude into a genuine rival to ChatGPT while making AI safety research part of its public commercial pitch rather than a side project.
xAI
xAI's rapid rise and its February 2026 absorption into SpaceX is one of the clearest signals yet that frontier AI labs are consolidating into companies with the balance sheets to fund trillion-dollar compute build-outs.
Together AI
Together AI is one of the clearest bets that open-weight models need their own specialized infrastructure layer rather than running on general-purpose cloud, and its Aramco-led round shows sovereign capital moving directly into the compute layer of AI rather than only into model companies.
Stability AI
Stable Diffusion's 2022 open-weight release is the reason so much of today's AI image-generation tooling exists at all - and the company's own turbulent finances and 2024 leadership overhaul is a cautionary case study in what open distribution does and doesn't do for a company's own balance sheet.
Scale AI
Nearly every frontier model released since 2020 was trained in part on data Scale AI's workforce labeled or ranked, and Meta paying $14.3 billion for a minority stake - and hiring its founder outright - was one of the clearest signals yet of how much Big Tech values control over AI training data.
Runway
Runway pushed generative video from a research curiosity into a tool film and advertising studios actually use, and its shift toward "world models" signals where the company thinks generative media is headed next - simulation, not just clip generation.
Presight AI
Presight is the one piece of G42's empire the public can actually see the books of - a listed, quarterly-reporting company that turns the UAE's sovereign AI ambitions into an audited revenue line, and one of very few AI-native companies of its kind listed on any exchange in the region.
Perplexity AI
Perplexity bet that people would rather read a synthesized, cited answer than click through a page of blue links, and grew from a $121 million valuation to roughly $23 billion in under three years betting on it.
OpenAI
ChatGPT is the product that took generative AI from a research demo to something hundreds of millions of people use weekly, and OpenAI's release cadence has set the industry's pace for most of the years since.
Mistral AI
The only European foundation-model lab operating at a scale comparable to the major US and Chinese labs, and the company most often cited when European governments talk about AI sovereignty that doesn't run through Silicon Valley.
Meta Superintelligence Labs
The clearest example of a Big Tech incumbent restructuring around the foundation-model race in real time - built by pulling in a rival's founder, and quietly abandoning the open-weight strategy that made Llama famous.
Hugging Face
Almost every open-weight model released by a foundation-model lab - Llama, Mistral, DeepSeek included - gets distributed through Hugging Face's Hub, making it the closest thing the AI industry has to shared infrastructure, and Nvidia's 2026 acquisition puts that infrastructure inside the company that supplies most of the industry's chips.
Google DeepMind
The lab behind AlphaFold and Gemini, and the clearest example of how the foundation-model race reorganised even Alphabet's own AI operation - the 2023 Brain/DeepMind merger, and the August 2026 leadership reshuffle under pressure from OpenAI and Anthropic.
G42
G42 is the anchor of the UAE's bet that a small, oil-wealthy nation can become a genuine AI infrastructure power - through direct partnerships with OpenAI, Microsoft, Nvidia and, after 2024, the US government itself - rather than only a customer of AI built elsewhere.
ElevenLabs
A huge share of the synthetic voices you now hear in AI products - audiobooks, dubbed video, voice assistants - were generated by ElevenLabs' models, and its rapid re-rating to an $11 billion valuation reflects how central voice has become to the AI product stack.
DeepSeek
DeepSeek's R1 model forced every major AI lab to publicly defend its compute spending in early 2025, and remains the clearest evidence that frontier-level AI research is no longer a US-only contest.
Databricks
Built on the open-source project that made large-scale data processing practical, Databricks has repositioned itself as the platform enterprises use to actually build AI on top of their own data - a less visible but arguably more durable business than a consumer chatbot.
Core42
Core42 is the physical layer under the UAE's AI ambitions - the data centres, GPUs and sovereign-cloud contracts that make "AI made in the UAE" a real infrastructure claim rather than a slogan, and increasingly a US-based infrastructure operator in its own right.
Cohere
Founded by a co-author of the paper that made modern AI possible, Cohere bet on enterprise and government customers over consumer chat - a narrower, less visible strategy than OpenAI's or Anthropic's, aimed at organisations that need models running on their own infrastructure.
Anthropic
Founded by researchers who left OpenAI over how it was approaching safety, Anthropic built Claude into a genuine rival to ChatGPT while making AI safety research part of its public commercial pitch rather than a side project.
xAI
xAI's rapid rise and its February 2026 absorption into SpaceX is one of the clearest signals yet that frontier AI labs are consolidating into companies with the balance sheets to fund trillion-dollar compute build-outs.
Together AI
Together AI is one of the clearest bets that open-weight models need their own specialized infrastructure layer rather than running on general-purpose cloud, and its Aramco-led round shows sovereign capital moving directly into the compute layer of AI rather than only into model companies.
Looking for one company?Browse all companies