2025

March

5 events

Editors' summary

Google takes the lead, OpenAI's image generation spills into the street

On 25 March, Google released Gemini 2.5 Pro, topping rival flagship models on the main evaluations and taking the lead.

The same day, OpenAI put GPT-4o's native image generation into ChatGPT. It spread as image generation that could render text without mangling it, and redrawing your own photographs in the style of Studio Ghibli became a worldwide pastime. Sam Altman changed his profile picture on X to one of them and wrote that the GPUs were "melting." An argument about how far a style may be imitated ran alongside it.

On the 31st, OpenAI raised $40 billion in a SoftBank-led round at a $300 billion valuation. On the 3rd, Anthropic had raised $3.5 billion in its Series E at a $61.5 billion valuation.

This block is written by the editors. It is kept separate from the sourced record below.

Record5 events
  1. 3
    BusinessAnthropic / Dario Amodei / Lightspeed Venture Partners / Bessemer Venture Partners

    Anthropic raises $3.5 billion at a $61.5 billion valuation

    Anthropic announced a $3.5 billion Series E at a $61.5 billion post-money valuation, led by Lightspeed Venture Partners with participation from Bessemer Venture Partners, Cisco Investments, D1 Capital Partners, Fidelity, General Catalyst, Jane Street, Menlo Ventures, and Salesforce Ventures. The company said the money would fund mechanistic interpretability and alignment research, expand its compute, and accelerate international growth. It came a week after the release of Claude 3.7 Sonnet and Claude Code, which were driving its commercial momentum.

  2. 14
    ProductGoogle

    Google says Gemini will replace Assistant

    Google said Gemini would replace Google Assistant on Android. Within the year the classic Assistant would no longer be accessible on most mobile devices or available for new downloads from app stores; tablets, cars, and phone-paired devices such as headphones and watches would move to Gemini, and a Gemini-powered experience would come to home devices like speakers, displays, and TVs. Devices below the minimum requirement of Android 10 and 2GB of RAM would keep the old Assistant unchanged. A product launched in 2016, which had made the give-it-a-command assistant familiar to a wide public, was being replaced by a conversational model. It followed the same line as Microsoft folding the Cortana app into Copilot in 2023: a generational handover among voice assistants happening in parallel across the industry. The migration did not keep to its schedule. On 19 December Google said completion would slip into 2026, with the mandatory transition on phones beginning 4 September 2026.

  3. 25
    ModelGoogle / Demis Hassabis

    Gemini 2.5 Pro puts Google back in front

    Google released Gemini 2.5 Pro as an experimental "thinking" model, the generation that built internal reasoning before answering into the default design, and it took the lead on benchmarks across mathematics, science, and coding. Gemini 2.5 Pro and Flash reached stable general availability on June 17. It was read at the time as the moment Google — on the back foot since ChatGPT — moved clearly ahead on capability comparisons.

  4. 25
    ModelOpenAI / Sam Altman

    Native GPT-4o image generation launches in ChatGPT

    OpenAI launched 4o Image Generation, image generation built natively into GPT-4o, making it the default image capability in ChatGPT. It rendered text without mangling it, followed instructions closely, and let you keep working on the same image through conversation; it went out to Free, Plus, Pro and Team users. Within days, redrawing your own photographs in the style of Studio Ghibli had become a worldwide pastime. Sam Altman changed his own profile picture on X to one of those images on the 27th, and posted that the company's GPUs were "melting" under the load. OpenAI said 130 million people made 700 million images in the first week. The company said it declined prompts naming the style of an individual living artist while permitting broader studio styles, a line that drew criticism. Hayao Miyazaki's remark that such work was "an insult to life itself" — made in 2016, watching a demonstration of CG animation — was quoted back widely, while Studio Ghibli itself said nothing about the trend.

  5. 31
    BusinessOpenAI / Sam Altman / Masayoshi Son / SoftBank Group

    OpenAI raises $40 billion led by SoftBank at $300 billion valuation

    OpenAI announced it would raise $40 billion in a funding round led by SoftBank, at a $300 billion post-money valuation, making it one of the largest private funding rounds ever at the time. OpenAI said the new capital would push the frontiers of AI research further, scale its compute infrastructure, and expand ChatGPT, which it said was used by 500 million people every week.