Generative AI Timeline

Anthropic publicly files an S-1 registration statement for a proposed IPO

OpenAI reports coordinated disclosure of critical PHP vulnerabilities discovered by AI agents

Google introduces a universal Gemini agent for work across enterprise tools and workflows

Mistral introduces Mistral Large 4

ElevenLabs introduces Eleven v4 and Eleven v4 Turbo text-to-speech models

AMD announces an agreement to acquire World Labs

OpenAI introduces GPT-6 Sol and GPT-6 Luna

OpenAI introduces the Agents API for building and deploying AI agents

Meta launches Muse, a personal AI agent that works across apps and services

OpenAI publishes a proposed AI-generated solution to the Navier-Stokes Millennium Prize Problem

OpenAI introduces GPT-6 Astra

Anthropic introduces Claude Fable 5.1 and Claude Mythos 5.1

Anthropic publishes results on Claude-led protein design and analytical chemistry, including experimentally tested protein binders

Stanford updates its observational study of AI exposure and employment among early-career workers

Munich court largely upholds GEMA copyright claims against AI music generator Suno

Google DeepMind introduces Gemini Robotics 2 for whole-body robot control

Black Forest Labs introduces FLUX 3, a multimodal model family for video, image and audio generation

US court grants final approval to the $1.5bn Anthropic copyright settlement over the acquisition and copying of books

OpenAI releases the GPT-5.6 model family

Anthropic releases Claude Sonnet 5

Anthropic launches Claude Fable 5 for general use and Claude Mythos 5 through restricted access

Anthropic raises $65bn in Series H funding at a $965bn valuation

Google introduces Gemini 3.5, a new model family built for agentic workflows

Google introduces Gemini Omni for generating video from text, images and video prompts

Microsoft and OpenAI announce an amended partnership agreement, expanding flexibility in cloud distribution

DeepSeek releases preview versions of its V4 open-weight models

OpenAI introduces ChatGPT Images 2.0

Stanford HAI publishes the 2026 AI Index Report

Anthropic announces Claude Mythos Preview and Project Glasswing to help secure critical software

Midjourney launches V8 in early-access alpha

Microsoft announces Copilot Cowork for delegating tasks across Microsoft 365, initially in preview

Anthropic publishes research on AI exposure and early evidence of labour-market impacts

Perplexity introduces Computer, a multi-model AI agent for completing complex projects

Anthropic releases Claude Sonnet 4.6

India hosts the AI Impact Summit in New Delhi

OpenAI begins testing advertising in ChatGPT in the US

Anthropic releases Claude Opus 4.6

OpenAI introduces Frontier, a platform for building and managing enterprise AI agents

OpenAI launches the Codex desktop app for managing multiple AI agents

Personal AI agent project adopts the OpenClaw name and announces a new release

Google introduces Personal Intelligence, connecting Gemini to information across Google apps

Anthropic introduces Cowork in research preview, bringing agentic capabilities to everyday knowledge work

OpenAI introduces ChatGPT Health, a dedicated experience for health and wellness

Boston Dynamics unveils the product version of Atlas and announces manufacturing and planned deployments

OpenAI releases GPT-5.2-Codex

Google release Gemini 3 Flash

OpenAI launch ChatGPT Image 1.5

Mistral introduce Mistral 3

Google launches Gemini 3

Adobe release Firefly Image 5

Anthropic introduce Agent Skills

OpenAI hosts its third DevDay, annoucing ChatGPT Apps, AgentKit, and Agent Builder

Anthropic introduce Claude Sonnet 4.5

OpenAI introduce ChatGPT Pulse

Anthropic pilots Claude for Chrome

Google DeepMind announce Genie 3 world model

OpenAI introduce ChatGPT Agent

Midjourney introduce V1 video model

Mistral introduce Mistral Code

OpenAI purchase Jony Ive's io design firm for $6.5bn

Dr. Jim Fan describes the Physical Turing Test for the first time

Amazon releases Nova 1.0 Premier, its most capable model so far

Google DeepMind announce support for Anthropic's Model Context Protocol

Measuring AI Ability to Complete Long Tasks research published

Google DeepMind release Gemini 2.5

OpenAI introduce next-generation audio models in their API

Google introduce AI Mode for Search

Amazon introduce Alexa+

Hugging Face announce AI Energy Score Ratings

OpenAI introduce Deep Research

International AI Safety Report published

DeepSeek release Reinforcement Learning research paper

DeepSeek release R1 model family under an open MIT licence

Titans: Learning to Memorize at Test Time research paper published by Google

Google DeepMind announces Veo 2 text-to-video model

OpenAI release GPT-4o advanced voice mode with video

Google DeepMind announces Gemini 2.0 Flash model

OpenAI launches '12 days of ship-mas' with the release of o1 and ChatGPT Pro subscriptions

Amazon announces Nova, its first family of multimodal frontier models

Stripe adds ability for LLMs to make payments

Apple releases iOS 18.1 with first Apple Intelligence features

John Hopfield and Geoffrey Hinton awarded the Nobel Prize in Physics for foundational discoveries and inventions that enable machine learning with artificial neural networks

Meta introduce Movie Gen text-to-video model

OpenAI hosts its second DevDay, annoucing the Realtime API, finetuning for vision, and prompt caching

Meta launch multimodal Llama 3.2

OpenAI launches o1 preview - a model that thinks

Adobe announces Firefly video model

Figure AI unveils its second generation humanoid robot

Mistral announces Mistral Large 2 Model

OpenAI launches GPT-4o Mini

Google announces Gemma 2, a family of lightweight open models

NVIDIA becomes world's most valuable company

Seoul Declaration published following the Global AI safety summit in Seoul

Ilya Sutskever leaves OpenAI

OpenAI announce GPT-4o

Google Deepmind introduce AlphaFold 3

Google releases Med-Gemini, a multimodal model specialised in medicine

Microsoft launches Phi-3, its smallest AI model yet

Boston Dynamic's Atlas robot goes fully electric

Udio launches v1 of its audio generation model

Stability.ai CEO resigns

NVIDIA announces its new Blackwell platform for the Generative AI era

Cognition.ai introduce Devin, the first AI software engineer

Figure AI raises $675m from Microsoft, OpenAI, NVIDIA and Jeff Bezos

Google DeepMind announces Genie that generates interactive environments

Stability.ai releases Stable Diffusion 3

Google releases Gemini 1.5 Pro just a week after releasing Gemini 1.0 Ultra

Google rebrands all of their AI features under the Gemini banner and released Gemini Ultra, their most capable model

Google announces that Chrome will be getting new GenAI features

Google's DeepMind unveiled AlphaGeometry that can solve geometry problems

OpenAI launches the GPT store that it announced at its DevDay in November

Google DeepMind introduce mobile ALOHA humanoid system

UK supreme court rules that AI cannot be named as a patent inventor

OpenAI inks deal with Axel Springer on licensing news for model training

Microsoft releases Phi-2

EU agrees the world's first laws to regulate AI

Google announces its long-awaited Gemini models

Pika Labs introduce Pika 1.0, their first text-to-video generation tool

Anthropic introduce Claude 2.1

Sam Altman fired as CEO of OpenAI

The Bletchley Declaration released following the UK hosted AI Safety Summit

Meta announces it will require political advertisers to disclose when images, videos or audio are altered with AI.

UK hosts the global AI safety summit

Google's AI chief says there's a 50% chance we will get AGI in just 5 years

Boston Dynamics shows off how it's been making Chat (ro)Bots

AI news readers get the spotlight. A virtual human called Zae-In spent five months reading live news bulletins on South Korean national broadcaster SBS

Adobe announces Fast Fill, generative fill for video, at the Adobe MAX conference

Snap's AI chatbot draws scrutiny in UK over kids privacy concerns

Character.AI in early talks for funding at more than $5bn valuation

Meta announces new AI experiences across their apps, including 28 AI characters played by major celebrities

OpenAI announces that ChatGPT can now see, hear and speak

Microsoft Copilot announced as an everyday AI companion

Amazon puts generative AI at the center of the latest Alexa upgrades

Harvard Business School and BCG release research into the effects of AI on knowledge worker productivity and quality

Anthropic introduce Claude Pro

Walmart announces it will give 50,000 office workers a generative AI app

OpenAI introduce ChatGPT Enterprise

Meta introduce the first all-in-one multilingual multimodal AI translation and transcription model

Google Chrome will summarize entire articles for you with built-in generative AI

The Blueprint launches with it's first weekly GenAI newsletter

US AI companies sign up to Whitehouse AI guidelines

Full episode of South Park reportedly AI-generated

Hollywood actors strike over AI threat

Anthropic releases Claude 2 chatbot, focuses on safety and alignment

ChatGPT outperforms 99% of Humans in Creativity Test

Black Mirror Netflix show paints creative but dystopian future for generative AI with Joan is Awful

EU Parliament adopts draft AI act, seeks to restrict use of generative AI

Generative AI company Cohere raises $270M in Series C, valued at $2bn

Marketers to save 5 hours/week by using generative AI, Salesforce study

AI discovers potential superbug-killing antibiotic

Generative AI startup Anthropic secures $450M in funding, valued at $1.8bn

OpenAI's Sam Altman testifies before US Senate on AI safety

EduTech Chegg stock crashes 48% on fears students migrating to ChatGPT

Film and TV screenwriters union (WGA) strike: Demand ban on AI-generated scripts

GenAI video of Will Smith eating spaghetti goes viral

American Medical Association: ChatGPT outperforms human doctors in both medical advice and patient empathy

Fake Drake x The Weeknd: Generative AI voice-cloned track raises concerns

AI-generated deepfake of Pope wearing Balenciaga puffer jacket goes viral

Adobe announces Firefly AI art generator

KitKat launches their AIa made this so we could take a break series

Midjourney's V5 AI image generator released, produces hyper-realistic synthetic photos

OpenAI releases GPT-4, a upgraded generative AI model with improved reasoning and creativity capabilities

Coca-Cola unveils Masterpiece ad created using generative AI

Meta unveils new open generative AI large language model LLaMA

Alphabet shares lost $100bn in value after a company event failed to impress and its new chatbot makes a mistake.

OpenAI introduce premium ChatGPT Plus subscription for $20/month

AI image generator Stability AI sued over copyright infringement

Mint Mobile releases advertisement scripted by ChatGPT

OpenAI launches ChatGPT, powered by GPT-3.5

Midjourney start testing v4 of their image generation model

Midjourney start testing v3 of their image generation model

MidJourney image generation platform launches on Discord

Midjourney starts beta testing v1 of their image generation model

Anthropic founded

Google Deepmind introduce AlphaFold 2

OpenAI release GPT-2 with 1.5bn parameters

OpenAI release GPT-1 with 117m parameters

Attention Is All You Need research paper published

Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift research paper published

Wait But Why publishes The AI Revolution: The Road to Superintelligence

Google acquires DeepMind for more than $500m

AlexNet wins the ImageNet 2012 challenge and becomes the first widely acknowledged, successful application of deep learning

DeepMind founded

ImageNet database released by Fei-Fei Li and team

NVIDIA release Compute Unified Device Architecture (CUDA)

A Neural Probabilistic Language Model research paper published

Speech recognition software developed by Dragon Systems becomes widely available

LeNet5, a convolutional neural network, proposed in a research paper by Yann LeCun et al.

Yann LeCun publishes Back Propagation Applied to Handwritten Zip Code Recognition research paper

Back propagation introduced in a research paper by David Rumelhart Geoffrey Hinton and Ronald Williams as a new learning feature for neural networks

Joseph Weizenbaum introduced ELIZA, the first computer that could communicate with humans by engaging in text-based conversations with rudimentary responses

Arthur Samuel introduced the term 'machine learning' with the first self-learning program created to play checkers on its own

John McCarthy proposes Artificial Intelligence as a new academic discipline at the Dartmouth Summer Research Project workshop

Alan Turing first mentions 'intelligent machinery' in a report exploring whether a machine could detect rational behaviour