OpenAI's charter
On 9 April 2018 OpenAI published a charter with four principles: broadly distributed benefits, long-term safety, technical leadership and cooperative orientation. It defines AGI as highly autonomous systems that outperform humans at most economically valuable work and names humanity as the organisation's primary fiduciary duty.
Why it matters
A laboratory wrote down as a commitment what until then had been voiced in statements about risk: if a value-aligned, safety-conscious project comes close to AGI first, OpenAI will stop competing with it and start assisting, a typical trigger being 'a better-than-even chance of success in the next two years'. A year later OpenAI LP names this charter as its primary obligation.
The charter also expects safety concerns to reduce OpenAI's traditional publishing while sharing safety, policy and standards research grows in importance, and holds that policy and safety advocacy without leading capabilities would be insufficient. The second source that the charter was applied is OpenAI's post on OpenAI LP in 2019: the new company's fiduciary obligation is to advance the charter's aims, and every investor and employee signs that it comes first. That is the same maker, and no independent confirmation was found, so confidence is medium. The charter's current address under openai.com/index answered 404; the archived copy of the 2018 blog was read.