The AI arms race

8th September 2025

With so much discussion in the tech world being focused on AI, who is ahead in the race for dominance in this new but rapidly expanding space?

A few years ago, AI was restricted to science fiction: a machine thinking for itself was pure fantasy. Then, when earlier models were introduced, they were impressive but rudimental – I remember using an early version of ChatGPT and it was quite limited in what it could achieve.

Now though, AI is all people can talk about. Many of us use it each day to help with repetitive tasks and processes, helping to automate and improve efficiency. It has grown considerably in power and capabilities, able of producing complex answers, or realistic videos that are not far off being as convincing as the real thing.

So, with all this discussion, and various AI companies fighting for consumer attention, who is on track to dominate the space? Who has been the most successful so far? And who has the most exciting plans for the future?

First, let us look purely at numbers – revenue often drives a company’s ability to innovate. OpenAI and Microsoft are the frontrunners, each with an estimated annual revenue of over $10 billion. Largely, this is due to how they are used commercially: for OpenAI, ChatGPT is seen as the household name for AI, it was the first one that people paid attention to, so they’ve sold a lot of subscriptions; for Microsoft, they already had the infrastructure to include Copilot in their existing products, so they had no trouble selling it forward. It makes sense that these two are dominating as Microsoft uses OpenAI models for Copilot: they each benefit from each other.

Anthropic and Google aren’t too far behind, with Anthropic growing fast. Meta, however, has made minimal revenue from AI as most of their access if free and open source. Additionally, little is known about Chinese corporation DeepSeek, as they haven’t shared revenue details.

In terms of current capabilities, each AI has its own strengths. OpenAI’s ChatGPT is capable of processing multiple data types and responds in real time, while Anthropic’s Claude is strong in reasoning, with a focus on safety, meeting high compliance benchmarks. Gemini from Google is best for long-context tasks, whereas DeepSeek is the most cost efficient. Each AI has works similarly, but has unique specialities, and it is difficult currently to see which will prove the most dominant.

Looking at some of the criticisms of AI, Open AI, Google, Anthropic and Meta have all warned that AI models are becoming too complex to monitor, potentially leading to safety issues down the line across the industry. Anthropic is dependent on only a few customers and Meta’s AI revenue is much lower than competitors, with no clear strategy in place.

Going forward, there are plans designed to fix some of these issues. Meta is betting on open source being profitable, but is also hoping to embed AI over its social media platforms, which would improve their reach significantly. Anthropic are looking to expand their partnerships, reducing their reliance on such a small part of the market.

Microsoft is planning to keep using OpenAI for model development, expanding what Copilot can do as OpenAI progresses. Google and DeepSeek are also both focusing on making general improvements and making their AI’s more accessible.

So, who is ahead and who is likely to dominate going forward? It is hard to say: we’ve not even looked at every AI model out there. Seemingly though, the partnership between OpenAI and Microsoft is one that, if it continues to be harmonious, could be the dominant AI force going forward. The balance of OpenAI being a reliable AI model, combined with Microsoft’s influence over the tech space, means that together, they’re unstoppable.

Do you agree, or do you think another company will win the AI arms race? Let Interfuture Systems know by contacting us HERE.

YouTube: https://youtu.be/PM5C5dDh9PI