OpenAI removes 'safely' from mission statement
OpenAI has quietly edited its core mission statement, removing the word “safely” from its objective of ensuring artificial general intelligence benefits all of humanity. The change is being interpreted as a pivot toward commercial priorities over long-term alignment concerns, testing whether the company can serve both shareholders and societal interests as it deploys more powerful AI systems.
- The previous mission statement, noted in 2022 and 2023 IRS filings, was "to build general-purpose artificial intelligence (AI) that safely benefits humanity, unconstrained by a need to generate financial return.” The new version, from the 2024 filing, was shortened to "to ensure that artificial general intelligence benefits all of humanity.” - This change follows an October 2025 corporate restructuring where OpenAI transitioned from a non-profit with a "capped-profit" arm to a for-profit Public Benefit Corporation (PBC) controlled by the non-profit OpenAI Foundation. This new structure removes the previous cap on investor profits. - In the new structure, the OpenAI Foundation holds a 26% stake, Microsoft holds 27%, and employees with other investors hold the remaining 47%. Despite holding only a minority equity stake, the non-profit Foundation maintains control over the for-profit PBC and can appoint and remove its board members. - OpenAI first introduced a for-profit subsidiary in 2019 to raise capital, creating a "capped-profit" model where investor returns were initially limited to 100 times their investment. This was a key mechanism for securing over $13 billion from Microsoft. - The mission statement has been edited multiple times since 2016. A 2018 revision removed a commitment to "openly share our plans and capabilities," and a 2020 change removed the phrase "as a whole" from "benefit humanity as a whole." - The company is currently facing several lawsuits concerning the safety of its products, with plaintiffs alleging psychological manipulation and wrongful death. - Competitors are using different approaches to governance and safety. Anthropic, founded by former OpenAI employees reportedly due to safety concerns, also operates as a Public Benefit Corporation with safety as a core part of its charter.
