Xi Jinping's First WAIC Keynote Comes With a New 29-Nation AI Body, Minus the US

The 2026 World AI Conference closed in Shanghai on July 20 after four days that China’s government is already calling a turning point for its AI diplomacy. President Xi Jinping opened the event on July 17 with his first-ever keynote at WAIC, an annual gathering that has run since 2018 without him. A day before that speech, 29 countries signed the founding agreement for a new body called the World Artificial Intelligence Cooperation Organization (WAICO), headquartered in Shanghai.

Neither the United States, the European Union, the UK, Japan, nor South Korea signed on.

TL;DR

WAIC 2026 ran July 17 to 20 in Shanghai and drew representatives from 102 countries, more than 400,000 on-site visitors, and over 3 billion online views according to the Shanghai municipal government. Xi Jinping attended the opening ceremony in person for the first time in the conference’s eight-year history and delivered a keynote calling for AI development to be a “symphony of international cooperation” rather than “a solo performance” by one country. The same week, 29 countries including Russia, Brazil, Belarus, Cuba, Venezuela, Indonesia, Pakistan, Kazakhstan, Laos, Cambodia, Malaysia, and Myanmar signed the agreement establishing WAICO, a new intergovernmental organization meant to coordinate AI standards, capacity-building, and governance among member states. China pledged 5,000 AI training slots for developing countries over the next five years. Huawei and ZTE used the show floor to debut hardware built to run without US components, including Huawei’s Atlas 950 SuperPoD and ZTE’s Nubia NaviX Ultra, billed as the first mass-produced AI agent phone. No G7 country joined WAICO as a founding member, and the launch lands squarely inside the wider contest between Washington’s export-control approach to AI and Beijing’s push to present itself as the open, inclusive alternative.

What happened at WAIC 2026

The conference ran under the banner “2026 World AI Conference and High-Level Meeting on Global AI Governance,” a pairing that signals how deliberately Beijing tied the trade-show side of the event to the diplomatic one this year.

Scale, by the numbers reported by the Shanghai municipal government:

• 102 countries and international organizations represented
• Over 400,000 on-site visitors across the four days
• More than 3 billion views of the event online worldwide
• 1,568 experts speaking, including 432 from outside China
• 11 Turing Award, Nobel Prize, or Fields Medal laureates on the program
• 158 forums and pitch sessions for 180 startups
• More than 300 AI products making their global debut
• Roughly 20.36 billion yuan (about $2.8 billion) in intended purchases signed at the event, up 25% year over year

Xi’s keynote on July 17 was the headline moment. He had never personally addressed WAIC before, sending video remarks or delegating the opening in prior years. This time he stood on stage and made the case that AI governance needs to be multilateral rather than dictated by whichever country gets furthest ahead first. He warned against any nation putting “its own security interests above those of others” and said the technology should be “for the positive, for good, and for humanity.”

On the industrial side, Huawei unveiled the Atlas 950 SuperPoD, a computing cluster linking 8,192 of its Ascend 950DT AI processors over a proprietary interconnect, built specifically to run without US-origin chips. ZTE’s Nubia brand, working with ByteDance’s Doubao assistant, launched the NaviX Ultra, which it calls the first mass-produced AI agent phone. Nubia says the 30,000-unit initial run sold out within hours of the announcement. Shanghai’s mayor also used the week to open a new Institute for Physical AI and Robotics, and organizers launched a 2026 International Embodied Intelligence Competition.

The World AI Cooperation Organization

The bigger story, in diplomatic terms, happened a day ahead of Xi’s keynote. On July 16, representatives from 29 countries signed the agreement creating WAICO in a ceremony in Shanghai.

WAICO is structured as an independent intergovernmental organization, not a UN body or a trade bloc, with its headquarters in Shanghai. Its stated purpose is to strengthen international cooperation on AI governance and support AI development that reporting describes as beneficial, safe, and fair.

The founding members named in Chinese state media and international coverage include Russia, Belarus, Serbia, Cuba, Brazil, Venezuela, Kazakhstan, Pakistan, Indonesia, Laos, Cambodia, Malaysia, and Myanmar, alongside roughly ten additional African nations not individually confirmed in the reporting available at publication. Full official membership rosters had not been independently verified beyond this list as of this writing.

Xi called the new body “a milestone in the history of world AI development” and framed it as an answer to Global South countries asking for more say in how AI gets governed globally. China has also pledged the 5,000 training and seminar slots for developing-country AI professionals over the coming five years, positioning the offer as part of WAICO’s practical mandate rather than a side gesture.

What stands out in the founding list is who is missing. No G7 country signed on. The US, the EU and its member states, the UK, Japan, and South Korea are all absent from the founding roster, a gap multiple outlets flagged within a day of the signing.

Why this matters

Two AI governance models are now running in parallel, and WAIC 2026 is the clearest marker yet of how far apart they’ve drifted.

Washington’s approach, laid out in the White House’s AI Action Plan released in 2025, centers on accelerating domestic AI infrastructure and tying access to advanced US AI systems to security vetting and export controls. Its international component is narrower: countering Chinese influence inside existing multilateral governance bodies rather than building a new one.

Beijing is doing the opposite. It’s building a new institution from scratch, open by design to countries that Washington’s export-control regime tends to restrict, and pairing it with an open-source posture on its own models. Xi’s “symphony, not solo” framing is a direct pitch to countries that feel shut out of the US-led AI conversation, and the 5,000 training slots make that pitch concrete rather than rhetorical.

It’s worth being precise about scale here. Twenty-nine countries is not a small number, but it’s also not the G20, and none of the world’s other major AI labs’ home countries signed on. Reuters and other outlets that covered the signing were careful to note the US, EU, UK, and Japan absence in the same breath as the announcement itself. Whether WAICO becomes a functioning standards body or stays a diplomatic gesture depends on what it actually does in its first year, not on the signing ceremony.

For the wider AI industry, the split matters beyond geopolitics. Two governance tracks moving in different directions means two sets of expectations for model transparency, safety testing, and data rules, and companies operating globally will increasingly have to build for both. That’s a governance story first. Its downstream effect on where AI research, capital, and talent concentrate is the part worth watching over the next year.

Frequently asked questions

When did the 2026 World AI Conference take place? WAIC 2026 ran July 17 to 20 in Shanghai. It was formally paired with a “High-Level Meeting on Global AI Governance” held during the same window.

Was this Xi Jinping’s first time speaking at WAIC? Yes. Xi attended the opening ceremony and delivered the keynote in person on July 17, 2026, the first time he has done so since the conference launched in 2018.

What is the World Artificial Intelligence Cooperation Organization? WAICO is a new intergovernmental organization founded by 29 countries on July 16, 2026, headquartered in Shanghai. Its stated mission is to coordinate international cooperation on AI development and governance.

Did the United States join WAICO? No. The US, along with the EU and its member states, the UK, Japan, and South Korea, did not sign on as founding members.

What products were announced at WAIC 2026? Huawei debuted its Atlas 950 SuperPoD AI computing cluster, built to operate without US-origin components. ZTE’s Nubia brand launched the NaviX Ultra, described as the first mass-produced AI agent phone, which reportedly sold out its initial 30,000-unit run within hours.

Primary sources and further reading

• Xi Jinping to attend World AI Conference for first time as China elevates tech push - South China Morning Post on Xi’s first in-person WAIC appearance
• Full text: Keynote speech by Chinese President Xi Jinping at opening ceremony of 2026 World AI Conference - Xinhua’s official transcript
• Top takeaways from Xi Jinping’s opening address at the World AI Conference in Shanghai - SCMP analysis of the keynote
• 29 countries sign agreement on establishing World AI Cooperation Organization - Xinhua on the WAICO founding signatories
• World Artificial Intelligence Cooperation Organization - Background and structure of the new body
• China creates World AI body with Russia and 27 others, but without US - CIO on the founding member list and G7 absence
• WAIC 2026 seals major deals, deepens global ties - Shanghai municipal government’s official closing recap and figures
• WAIC ends in Shanghai on a high note - China Daily on the conference’s closing outcomes
• Huawei’s new computing cluster, world’s first AI agent phone to debut at China AI summit - SCMP on the Huawei Atlas 950 and ZTE NaviX Ultra launches
• Xi Jinping positions China as open-source AI leader at WAIC 2026 - Quartz on the open-source governance framing
• Reading between the lines of the dueling US and Chinese AI action plans - Atlantic Council comparison of the two countries’ AI policy approaches