A Salesforce CEO Sat Next to a President — and the UN Has Never Drawn This Picture Before

On July 8, in Geneva, a genuinely strange tableau took shape. The AI for Good Global Commission — a 44-member body launched by the UN's International Telecommunication Union (ITU) — convened for the very first time. And the people around that table made for an odd portrait. On one side you had heads of state: Rwandan President Paul Kagame, Estonian President Alar Karis, Icelandic President Halla Tómasdóttir. Right beside them sat corporate chiefs: Nvidia CEO Jensen Huang, Amazon CEO Andy Jassy, Microsoft President Brad Smith.

Here's why that's the news. Until now, the UN has essentially never seated a corporate CEO as a full, equal member of an official governance body alongside heads of state. Companies usually advised from the back — as observers, or through consultative panels. This time, Salesforce CEO Marc Benioff isn't a bystander. He is co-chair of the commission, sharing the gavel with President Kagame. A sitting president and a public-company CEO holding equal authority over a UN-badged governance forum is, plainly, unprecedented.

The timing is almost too neat. Just one day earlier, on July 7, in the same city, the UN's first Global Dialogue on AI Governance wrapped up (the meeting we covered yesterday — the one where Secretary-General Guterres demanded a killer-robot ban and 169 countries showed up). The moment that closed, this commission broke ground the next morning, folded inside the ITU's AI for Good Summit (July 7–10). In the span of a single day, the stage shifted from a government-led dialogue to a hybrid corporate-plus-government commission. It was a week that made Geneva's ambition — to become the permanent capital of international AI rule-making — impossible to miss.

Who's on the Stage — the ITU, Two Co-Chairs, and a Big-Tech Battalion

Start with the convener: the ITU. The International Telecommunication Union, founded in 1865, is the oldest specialized agency in the UN system. It began as a body for telegraph, telephone and spectrum standards, but in the digital era it has been aggressively repositioning itself as the venue for AI governance. Since 2017 it has run the AI for Good multi-stakeholder platform, hosting an annual summit in Geneva. Secretary-General Doreen Bogdan-Martin serves as the commission's permanent vice-chair, and her line captures the whole spirit of it: "No organization can single-handedly put AI at the service of all humanity. It will take collective leadership and the combined expertise of partners across sectors."

The pairing of the two co-chairs is itself the message. Rwandan President Paul Kagame is the face of the Global South. Rwanda has invested aggressively in digital infrastructure across Africa, and Kagame has personally pushed AI as a national-development strategy. His framing set the tone: "Technology is supposed to be a force for good, and we have a responsibility to use it accordingly." For a developing country, AI isn't an abstract existential risk — it's a question of whether you get to be in the room at all.

Salesforce CEO Marc Benioff anchors the opposite pole. Salesforce sells AI products (Agentforce and the rest) and books revenue in nearly every country represented on this commission. So the structure is this: the "seat where the rules get written" is co-chaired by someone whose company will be subject to those rules and is chasing those markets. That's precisely why the pairing draws two opposite readings at once — a symbol of public-private cooperation to some, a fox-guarding-the-henhouse arrangement to others.

The member roster only sharpens that tension. It includes Nvidia's Jensen Huang, who all but monopolizes AI chips; Amazon's Andy Jassy, who runs the world's largest cloud; Microsoft President Brad Smith; Anthropic co-founder Jack Clark, a leading voice on AI safety; Cohere co-founder Aidan Gomez; Google/Alphabet's James Manyika; and Qualcomm CEO Cristiano Amon. Indian industrial titans Mukesh Ambani (Reliance) and Sunil Bharti Mittal (Bharti) joined too. On the government side, beyond the Rwandan, Estonian and Icelandic presidents, AI and tech policymakers from Kazakhstan, Namibia, Nigeria, Singapore, Togo and Saudi Arabia signed on. It is, quite literally, "the people who build AI + the people who'd regulate it + the people who desperately need it" crammed into one room.

What It Actually Intends to Do — the First Agenda and the Number 44

The focus of the first meeting (July 8) was surprisingly concrete. Not killer robots, not AGI doom — but AI infrastructure and its applications in health, education, food security and disaster response. In other words, the emphasis landed less on "what should we regulate" and more on "how do we solve the real problems of the countries AI has left behind." The commission's stated top priority is equally clear: closing the digital divide facing the 2.2 billion people who still aren't even online.

The official mission boils down to three phrases: expand access, strengthen trust, accelerate impact. And a fourth commitment underneath them — to make developing countries active shapers of the global AI agenda rather than passive recipients of it. That's the whole reason Kagame is co-chair. Right now the compute, data and capital of AI are concentrated in a handful of companies in the US and China, and Africa, Central Asia and Latin America are effectively in the audience.

Here it is by the numbers.

Item Detail
Body AI for Good Global Commission
Convener UN / ITU (International Telecommunication Union)
Launch announced July 2, 2026
First meeting July 8, 2026, Geneva
Size 44 founding members
Co-chairs Marc Benioff (Salesforce CEO) + Paul Kagame (President of Rwanda)
Permanent vice-chair Doreen Bogdan-Martin (ITU Secretary-General)
Corporate members Jensen Huang (Nvidia), Andy Jassy (Amazon), Brad Smith (MS), Jack Clark (Anthropic), Aidan Gomez (Cohere), and more
Government members Presidents of Rwanda, Estonia, Iceland + policymakers from Kazakhstan, Namibia, Nigeria, Singapore, Togo, Saudi Arabia
First-meeting agenda AI infrastructure; health, education, food security, disaster response
Core goal Close the 2.2-billion-person digital divide; access, trust, impact
Nature Advisory, non-binding multi-stakeholder platform

One thing to state coldly here: this commission does not make law. It has no authority to issue binding treaties or regulations. That's exactly why an outlet like Tech Times went out of its way to headline a piece on "what this commission can and cannot do." Officially, it's a multi-stakeholder platform for advice, recommendation and agenda-setting — not an enforcement mechanism. Which means if the consensus reached here doesn't translate into national legislation, it can end as "great photos and a press release."

What Each Side Gets Out of It

For the ITU and the UN, the commission is a play to cement Geneva as the capital of AI norms. If the intergovernmental dialogue that ended a day earlier was "a stage for governments," this commission is "a hybrid stage for governments plus companies," which lets it sketch a far more actionable picture. The 2.2-billion digital-divide framing dovetails cleanly with the UN's Sustainable Development Goals, giving it airtight justification. For the UN, it's a showcase that multilateralism still functions in the AI era.

For the corporate CEOs, the upside is far more naked. First, sitting in the room where rules get written is a chance to tune those rules toward your own business. Second, the banner of "AI access for developing countries" is, functionally, a new market. Nvidia sells chips, Amazon and Microsoft sell cloud, Salesforce sells software — and a customer base of 2.2 billion people is exactly where "doing good" and "expanding your market" overlap perfectly. Third, reputation. In an era of intensifying regulatory pressure, the title "UN commission member" is a powerful shield.

For developing countries, a channel to be heard has opened. Until now, AI-norm debates were led by the US, EU and China, with Africa and Central Asia informed of decisions after the fact. The symbolism of Kagame as co-chair, and of countries like Namibia, Togo and Kazakhstan being founding members, means at minimum that "we were in the room when the board was set." Whether actual resources — chips, cloud, education — flow in is a separate question.

Conversely, civil society and critics aren't dropping their guard. Outlets like Common Dreams flatly called the commission a body "full of Big Tech execs." Isn't it a conflict of interest, they ask, for the very entities that ought to be regulated to co-chair the table designing the regulation? And because the body is non-binding, its real checks are thin — a point critics keep returning to. Beneath it all sits the worry that companies are laundering self-serving norms through a borrowed UN logo.

Past Parallels — the Wins and the Flops

This isn't the first experiment in "hybrid public-private global governance." A frequently cited success is ICANN, which manages the internet's addressing system. Its multi-stakeholder model — governments, companies, the technical community and civil society all participating — has effectively run the world's domain system without any single government owning it. As proof that seating diverse stakeholders together can produce governance that actually works, it's a model this commission could learn from. Getting all the interested parties around one table can itself be a source of strength.

But there are painful failures too. Think of the many public-private initiatives championed by the World Economic Forum at Davos. They launched with glittering CEO lineups and grand declarations, yet lacking any binding force, many never translated into real change and earned the jibe of "photos at Davos, then nothing." The lesson is unambiguous: gathering famous people and actually changing something are entirely different problems. That's precisely the trap observers warn this commission risks falling into.

One more worth referencing is GAVI, the vaccine alliance in the health sector. A public-private partnership built by governments, pharmaceutical companies, the Gates Foundation and the WHO, it produced concrete results — actually delivering vaccines to developing countries. The takeaway there: when money and an execution mechanism are attached, a hybrid model can deliver real outcomes. Whether the AI for Good Commission builds a GAVI-style fund and distribution structure, or ends up a Davos-style declaration machine, is the fork in the road that decides its fate.

How Rivals Play It

This commission isn't the only player trying to write AI's rules. The furthest along is still the EU. It has already brought its AI Act into force, running risk-tiered regulation, and it carries the confidence of "we already have binding law." From Brussels' vantage point, a UN commission can look like a nice-but-toothless advisory body — because real regulatory power flows from the law of a large market, which is the logic of the "Brussels effect." Expect the EU to treat this commission mostly as a megaphone for exporting its own standards globally.

The US reads differently again. Its instinct prioritizes innovation and industrial primacy, which makes it cautious about international binding force that would tie the hands of its own tech giants. Yet this commission is stacked with US corporate CEOs — Huang, Jassy, Smith, Clark. That signals a US preference for "flexible, self-directed governance with companies directly participating" over "a regulatory treaty." In fact, Washington may welcome the commission's non-binding character as exactly the right amount of governance.

China occupies an awkward spot on this board. Chinese Big Tech (Baidu, Alibaba, Tencent) and Chinese government representatives are conspicuously absent from the founding roster. On a commission built around US companies, China is likely to keep its distance — and instead push an alternative narrative through BRICS or its own multilateral channels: that "we are the true voice of the developing world." How China counters the Global South credibility that a Kagame co-chairmanship lends this commission is one of the story's live questions.

And don't forget the AI companies' own double game. Firms like OpenAI, Google and Anthropic have led with "self-regulation" and "voluntary safety pledges," trying to seize the initiative before hard government regulation arrives. Joining this commission is an extension of that same strategy: get to the regulatory table early so the rules don't drift against you. The inclusion of figures like Anthropic's Jack Clark and Cohere's Aidan Gomez — one symbolizing "safety," the other "openness" — looks like a calculated placement.

So What Actually Changes

For everyday users, there's almost no change you'll feel right away. The chatbot you used today won't be different tomorrow. But the direction is worth reading. The fact that this commission's center of gravity sits on "expanding access" rather than "regulation" signals that AI policy discourse over the next few years may shift from "how do we contain AI" toward "how do we spread AI to more people." If you live in a developing region or an underserved area, that potentially opens a channel to become a beneficiary of AI education and infrastructure programs.

If you work in AI or run a startup, read the composition itself as a signal. Big-Tech CEOs sitting as co-chairs and members of a UN commission means future global AI norms are likely to be shaped "with companies participating and tuning them." That tends to favor large incumbents and disadvantage smaller players — a startup that can't get into the rule-design room ends up on the receiving end of decisions. On the flip side, as the "AI access for developing countries" market opens up, there may be real opportunity for shops building low-cost, low-power AI solutions.

For anyone tracking policy or investment, this is a meaningful milestone. It shows the center of gravity in AI governance expanding from "intergovernmental treaties" (yesterday's Global Dialogue) to "hybrid public-private commissions." The fact that both tracks ran back-to-back in Geneva in the same week signals that the city wants to become AI's permanent standards hub — the way the IPCC is for climate or the WTO is for trade. How companies like Salesforce, Nvidia and Microsoft position themselves here ties directly to future regulatory risk and market access, which makes it a trend investors should watch.

A note of sobriety, though: this commission is still a forum for dialogue, not a body of law. If the glittering roster and grand mission don't translate into actual resource allocation and national legislation, it risks settling into a Davos-style photo-op institution. The real test comes at the second and third meetings — whether concrete funds and execution mechanisms actually emerge.

🥄 Three Things You're Probably Wondering

— So what does this mean for me? Not much right now. The AI services you use won't change tomorrow because of this meeting. But since the commission weights "getting AI to more people" over "regulation," it could shape the direction of AI education and infrastructure programs a few years out. If you live in a developing country, that's a more direct signal.

— Is it okay for corporate CEOs to be writing the rules? It's contested. Supporters argue the people who understand AI best should help write the rules, or the rules won't be realistic. Critics call it a conflict of interest — the regulated designing the regulation. Common Dreams bluntly labeled it a commission "full of Big Tech." And because the body is non-binding, its checks are weak, so we'll have to watch whether that worry becomes real.

— How's this different from the UN meeting that ended yesterday? Different layer. Yesterday's Global Dialogue (which closed July 7) was a government-led stage where 169 countries debated regulatory issues like killer robots and child safety. Today's commission is a 44-member hybrid of corporate CEOs and heads of state, focused on "access and development" more than "regulation." Running both in Geneva a day apart looks like a UN strategy to push regulation and diffusion along two parallel tracks at once.

Sources

Numbers are as of announcement and may change.