AI4COP at CBD COP17

AI now shapes how Parties prepare their positions, how delegations read draft texts, and how decisions are carried out at home. Most of these tools are built and hosted by a few companies in a few countries. Who controls the sensitive data? Whose values do the models carry? Can every country use AI on its own terms?

H.E. Dr. Gevorg Mantashyan, Armenia's Deputy Minister of High-Tech Industry and Artificial Intelligence, opens the session with an address and then joins the panel.

At the session we launch PACE, four principles for sovereignty-aligned AI in environmental multilateralism: Participation, Accountability, Consent and Equity. Working groups drafted them at the AI4PG workshop at ETH Zurich in July. In Yerevan we open them for public comment.

In August, at UNCCD COP17 in Ulaanbaatar, a first dinner started a coalition of AI developers, negotiators, UN bodies and governments to agree shared principles for AI in multilateral diplomacy.

In Yerevan we sit down again, over food and informal conversation, to work through the values sovereignty-aligned AI should hold and the principles that should govern it. The evening follows the PACE launch two days earlier.

Notes from the first dinner are on the interest group page.