ChatGPT

OpenAI was founded in December 2015 in Delaware as a nonprofit research lab by Sam Altman, Greg Brockman, Ilya Sutskever, Elon Musk, and others, with a stated mission to develop artificial general intelligence (AGI) for the benefit of humanity. Musk departed the board in 2018 after disagreements, and in 2019 OpenAI created a capped-profit for-profit subsidiary to attract the enormous capital required for large-scale model training. Microsoft committed an initial $1 billion investment that year, eventually totaling over $13 billion across subsequent rounds. By 2020, OpenAI released GPT-3 — trained on 175 billion parameters and 45 terabytes of text data — which powered early developer applications via API. In January 2022, OpenAI released InstructGPT, a reinforcement-learning-from-human-feedback (RLHF) fine-tune of GPT-3.5 that became the direct technical precursor to ChatGPT. A 2025 restructuring converted the for-profit subsidiary into a public benefit corporation (OpenAI Group PBC), with the nonprofit OpenAI Foundation retaining 26% conventional equity.

ChatGPT launched on November 30, 2022 as a free public research preview built on GPT-3.5. OpenAI's own team released it internally described as 'a research preview' with few expectations; engineers later said they were surprised by how much mainstream attention it attracted. ChatGPT reached 1 million users in five days and 100 million monthly active users in roughly two months, making it the fastest-growing consumer internet application in history at the time. ChatGPT Plus, a $20/month subscription offering faster access and priority features, launched on February 1, 2023. On March 14, 2023, OpenAI released GPT-4, a substantially more capable multimodal model capable of accepting both text and image inputs, available first in ChatGPT Plus and via API. Microsoft rapidly integrated OpenAI models into Bing, Edge, and Microsoft 365. In March 2023, Italy became the first country to temporarily ban ChatGPT over data-protection concerns, lifting the block one month later after OpenAI addressed regulatory demands. The New York Times sued OpenAI and Microsoft in December 2023 for alleged copyright infringement, and the case entered the discovery phase.

In November 2023, OpenAI experienced the most dramatic corporate governance crisis in modern tech history. On November 17, 2023, the nonprofit board fired CEO Sam Altman, citing that he 'was not consistently candid in his communications with the board.' Former board member Helen Toner later alleged on a podcast that Altman had withheld information about safety processes and that two executives had reported 'psychological abuse.' Co-founder and president Greg Brockman resigned in solidarity the same day. Microsoft publicly announced plans to hire Altman for a new AI research division. Within hours, the vast majority of OpenAI's roughly 800 employees signed a letter threatening to resign and join Altman at Microsoft if the board did not reverse course. Facing the collapse of the company's workforce and investor relations, the board reinstated Altman as CEO on November 22, 2023 — just five days after his ouster — and several board members who voted to remove him subsequently departed.

In 2024, OpenAI expanded both its technical capabilities and its legal and regulatory exposure. GPT-4o ('omni') launched in May 2024 as the first model to natively process and generate text, audio, and images in a single model without pipeline handoffs, achieving state-of-the-art results in multilingual and voice benchmarks. A separate controversy erupted after users noted a resemblance between GPT-4o's 'Sky' voice and actress Scarlett Johansson's voice; OpenAI pulled the voice on May 20, 2024, and said it was not an imitation. On September 12, 2024, OpenAI launched o1-preview, the first model in its 'o-series' reasoning paradigm, which allocates additional inference-time compute to chain-of-thought thinking before responding, achieving approximately PhD-level performance on physics, chemistry, and biology benchmarks. The full o1 model and the $200/month ChatGPT Pro tier launched in December 2024. Italy's Garante concluded its 2023 investigation in December 2024 with a €15 million GDPR fine against OpenAI.

GPT-5 launched on August 7, 2025 as the default model for all ChatGPT users, offering approximately 45% fewer factual errors than GPT-4o on web-search-enabled prompts and new state-of-the-art results on math, coding, health, and visual reasoning benchmarks. OpenAI then entered an aggressive sub-major iteration cycle: GPT-5.1 (November 2025), GPT-5.2 with Thinking/Pro tiers (December 2025), GPT-5.3/GPT-5.4 with deepened Codex integration (March 2026), and GPT-5.5 — codename 'Spud' — on April 23, 2026, which became the current flagship model. GPT-5.5 Instant replaced GPT-5.3 as the free-tier default on May 5, 2026. During this period, OpenAI closed a $122 billion funding round in March 2026 at an $852 billion valuation, faced a wrongful death lawsuit in August 2025 alleging ChatGPT contributed to a minor's suicide, signed a classified-network deployment deal with the U.S. Department of Defense in February 2026, and by May 2026 was reported to be confidentially filing for an IPO targeting a fall 2026 debut.