If a new AI safety agreement between the U.S. and Chinese governments is formally announced after Issuance and reported by any of the Source Agencies before Dec 1, 2026, then the market resolves to Yes. An announcement must state that both governments have reached agreement and identify at least one specific agreed measure to assess, prevent, reduce, or respond to harm from AI systems or their misuse. The measure must specify an action, restriction, standard, procedure, or research or operational arrangement that one or both governments agree to undertake or support. A preliminary agreement, agreement in principle, or framework qualifies if it meets this test; signing, final text, ratification, legal enforceability, and implementation are not required.
A formal announcement must appear in an official agreement text, communiqué, readout, press release, or other written government statement announcing or confirming the agreement, or be made by an authorized government representative at a government press conference or briefing, leaders' summit, signing ceremony, or formal address. An announcement by either government that expressly states both governments have agreed is sufficient. Informal remarks, including press gaggles, do NOT qualify on their own; publishing their transcript does not change this. A later qualifying formal announcement counts at its own announcement time.
A qualifying measure may concern safety testing, incident reporting or response, safeguards against misuse, restrictions on dangerous development or uses, human control, or defined safety research. These examples do not limit the test above. Agreements limited to future discussions, negotiations, general cooperation, or broad aspirations for safe or responsible AI do NOT qualify. A statement that an agreement exists without identifying a specific agreed safety measure does NOT qualify.
A qualifying AI safety component of a broader agreement counts, including a multilateral agreement if both governments expressly adopt its qualifying commitments. Private-company or unofficial-dialogue agreements without government adoption do NOT qualify. The agreement must be newly announced during the market's window; a later report or reaffirmation of a previously announced agreement does NOT qualify. A newly announced amendment, extension, or replacement qualifies if it includes a specific agreed safety measure applying to a new scope or period.
If a new AI safety agreement between the U.S. and Chinese governments is formally announced after Issuance and reported by any of the Source Agencies before Dec 1, 2026, then the market resolves to Yes. An announcement must state that both governments have reached agreement and identify at least one specific agreed measure to assess, prevent, reduce, or respond to harm from AI systems or their misuse. The measure must specify an action, restriction, standard, procedure, or research or operational arrangement that one or both governments agree to undertake or support. A preliminary agreement, agreement in principle, or framework qualifies if it meets this test; signing, final text, ratification, legal enforceability, and implementation are not required.
A formal announcement must appear in an official agreement text, communiqué, readout, press release, or other written government statement announcing or confirming the agreement, or be made by an authorized government representative at a government press conference or briefing, leaders' summit, signing ceremony, or formal address. An announcement by either government that expressly states both governments have agreed is sufficient. Informal remarks, including press gaggles, do NOT qualify on their own; publishing their transcript does not change this. A later qualifying formal announcement counts at its own announcement time.
A qualifying measure may concern safety testing, incident reporting or response, safeguards against misuse, restrictions on dangerous development or uses, human control, or defined safety research. These examples do not limit the test above. Agreements limited to future discussions, negotiations, general cooperation, or broad aspirations for safe or responsible AI do NOT qualify. A statement that an agreement exists without identifying a specific agreed safety measure does NOT qualify.
A qualifying AI safety component of a broader agreement counts, including a multilateral agreement if both governments expressly adopt its qualifying commitments. Private-company or unofficial-dialogue agreements without government adoption do NOT qualify. The agreement must be newly announced during the market's window; a later report or reaffirmation of a previously announced agreement does NOT qualify. A newly announced amendment, extension, or replacement qualifies if it includes a specific agreed safety measure applying to a new scope or period.
No volume detected, be the first!
