Welcome to P3 Media’s AI Commerce Brief, your daily update on the AI and commerce stories shaping how companies build, sell, and grow. It’s Tuesday, July 28, 2026. Let’s get into it.
Our top story is infrastructure. Meta and BlackRock announced a new venture to develop and own a data center campus in El Paso, Texas. Meta says the campus is under construction, will have one gigawatt of compute capacity, and is expected to begin bringing capacity online in 2028.
The financing structure matters. Funds managed by BlackRock will own 80 percent of the venture, while Meta keeps 20 percent. Meta says the parties are funding their share of about $14 billion in development costs for buildings, power, cooling, and connectivity. Meta will lease the full campus and says the project supports its AI models and core business.
Next, NVIDIA and Safe Superintelligence, the lab led by Ilya Sutskever and Daniel Levy, announced a long-term strategic partnership. NVIDIA says it has made an investment in SSI and will give the company access to Vera Rubin systems. NVIDIA says that access will expand SSI’s compute by an order of magnitude.
The signal is clear. Frontier AI labs are competing not only on model quality, but on guaranteed infrastructure relationships.
Now to the open model debate. Anthropic CEO Dario Amodei wrote that Anthropic has never advocated for a category ban on open-weight models. He said open-weight models without dangerous capabilities can be a public good. But he also argued for keeping powerful chips out of authoritarian hands, cracking down on industrial-scale distillation, and requiring safety testing for sufficiently capable models, whether open or closed.
At nearly the same moment, NVIDIA announced the Open Secure AI Alliance, with partners across cloud, enterprise software, cybersecurity, and open source. NVIDIA says the alliance will develop and share open tools for AI safety and cybersecurity.
The tension here is worth watching. One side says open systems give defenders more control. The other says the most capable open systems need stronger testing before release. Both views are now shaping enterprise AI policy.
OpenAI also published new workplace research. In an analysis of more than 800,000 US ChatGPT messages, OpenAI says 43.5 percent of non-generic, occupation-specific messages involved tasks outside the user’s own occupation. The pattern was especially visible in customer experience, design, human resources, legal, and marketing.
For commerce teams, this is the practical AI story. Marketers troubleshoot websites. Salespeople analyze data. Small-business owners draft copy, review contracts, and run basic financial analysis. AI is not just automating tasks. It is changing who does the work.
In commerce pulse, ACI Worldwide and dLocal announced a partnership giving global merchants access to local payment methods in Brazil and Mexico through ACI’s payments orchestration platform, with four more Latin American markets planned. Separately, TSG announced a partnership with Stripe focused on helping platforms and software providers treat payments as a growth lever.
What to watch: whether Meta’s El Paso financing closes as expected, whether the open AI security alliance produces usable tools, and whether AI teams start measuring role redesign as carefully as they measure cost per token.
That’s your AI Commerce Brief for today. Thanks for listening.
