FAQs about OpenAI and the Proposed Effingham Development

Not only does OpenAI not make a profit at this point, its CEO has consistently been accused of saying anything that he can to get what he wants with little concern for the people around him. Consider what he thinks of communities in GA if he treats his coworkers this way.

But let’s just look at the finances of the company:

2024 — OpenAI Had $3.7 Billion In Revenue, $12.4 Billion In Costs and Expenses, and a net loss attributable to the company of $5.09 Billion.

2025 — OpenAI Had $13.07 Billion In Revenue, $34 Billion In Costs and Expenses, and $20.92 Billion In Losses, with a net loss attributable to the company of $38.53 Billion

Please note that 2025 was the year that OpenAI converted from a non-profit to a for-profit entity, leading to a $41.55 billion loss due to changes in fair value of convertible interests and warrant liability.

On July 22, 2026, TechCrunch noted that OpenAI is now claiming they will spend $750 billion dollars on data centers yet, Stargate data center has stalled because costs are soaring.

Fast Company reported in April that OpenAI doesn’t expect to be profitable until 2030, but the amount of debt they are incurring coupled with corporate pullback from LLMs like OpenAI because they see no return on investment (ROI) creates doubt that profits could materialize. Now, with opensource AI models that can run on your own servers or computer, companies are finding cheaper alternatives.

If this project is reliant on Oracle, OpenAI might have an issue finishing the project as we are seeing with Stargate. Oracle’s rating on the market is one step above a junk bond, why? Because it is stuck in a cash to debt cycle and there is no light at the end of that tunnel. Additionally, investors are suing Oracle for not disclosing debt connected to OpenAI to the tune of $56 billion.

QUESTION TO ASK: Can OpenAI afford a surety bond that could pay for decommissioning and restoration of 1400 acres?

Below is from Project Camellia FAQs and Documents page (which has no supporting documents). Under water use it says:

The project is planned to use a closed-loop water system that recirculates water rather than continuously withdrawing and discharging large additional volumes. It works like a car radiator and is designed to keep ongoing water use low. The initial water supply would be sourced in a way designed not to affect community water needs, and the project would not rely on local public water supplies for ongoing cooling. Once fully built, the project is expected to employ thousands of people. Ongoing water use will primarily support ordinary workplace needs, such as sinks, restrooms, kitchens, sanitation, and maintenance, and is expected to be comparable to that of other workplaces with a similar number of employees. OpenAI also plans to share water-usage data publicly.

Let’s highlight some things from that paragraph

1. “The initial water supply would be sourced in a way designed not to affect community water needs, and the project would not rely on local public water supplies for ongoing cooling.”
   Question to ask: Where are they getting the water from? And how much water will actually be used?
2. “OpenAI also plans to share water-usage data publicly.”
   Question to ask: How? Will they create a transparent resource use dashboard?

QUESTION TO ASK: Who are they contracted with for the architecture for this project? That will impact how much water will be used and how hot the surrounding area will become.

“While electricity from Georgia Power is expected to start flowing in 2028, OpenAI did not give a timeline for when the first GPU will be turned on. That could happen sooner than 2028 given that OpenAI recently hired Brett Mayo to lead data center construction. Mayo previously worked at xAI, where he oversaw the Colossus data center in Memphis.”(Techcrunch)

This facility is located close to proposed SSE4 gas pipeline expansion, namely the Rincon compressor station.

Question to ask: How can we guarantee that this will not be a behind-the-meter power station like the xAI in Memphis?

We can only speculate what the water use is. OpenAI may use Oracle hardware but as their February 9, 2026 article “In the simplest terms: The heat leaves the building; cooling liquid does not.” (Oracle) If OpenAI will be using this direct-to-chip, closed-loop, non-evaporative architecture, they claim “There is no evaporation, blowdown, or continuous makeup water requirement” but when the heat leaves the building, as it says, it will increase the temperature in the surrounding area. Yet in the section on their FAQs page they say the following. (bolding was by us)

“A: No, we do not expect the project to cause a noticeable change in temperatures in the surrounding community. Data center equipment is highly sensitive—and represents a substantial investment—so maintaining carefully controlled temperatures is essential to reliable operation. The facility’s cooling systems will be designed to manage and dissipate heat safely, with operating conditions continuously monitored. Simply put, effective temperature management is fundamental to protecting the equipment and keeping the facility operating properly.”

This is non-committal language. A 2026 Study showed that data centers create “local microclimate zones” or heat islands. The study found an average temperature change measured about 3.6 degrees F with highs of 16 degrees F above surrounding area temperatures.

QUESTION TO ASK: What will be done to measure the heat impacts and what funds are budgeted to address a heat island effect?

Maybe they will use the JetCool system by Flex who is working with Broadcom who entered into a $10 billion agreement with OpenAI while they still rely on Oracle’s Nvidia chips for other projects. This raises the same question about where the heat will go.

Probably – odds are “yes” as the AI hype is following a known economic cycle – the Gartner Hype Cycle – and most likely the AI boom is nearing the “Peak of Inflated Expectations.”

For those of you old enough to remember the start of the internet in the late 1990s-early 2000s — there was a huge amount of excitement, a lot of (over)investment, a recession bust and, now, the internet is a viable part of our economy. But that took a while.

News stories that track to the “bubble”

Google and Tesla shares plunge as AI spending rattles markets | BBC | 23 Jul, 2026

IBM Lowers Its Growth Outlook as Sales of Data-Center Mainframes Sink 42% | WSJ | 22 Jul, 2026

A.I. Spending Sets a Record, With No End in Sight | The New York Times | 29 Apr, 2026

People Loved the Dot-Com Boom. The A.I. Boom, Not So Much | The New York Times | 21 Feb, 2026

Adoption has plateaued. In the fourth quarter of 2025, 38 percent of employees told Gallup that their workplace had integrated A.I. technology. That number was essentially unchanged from the third quarter…A.I. is clearly not technology that is being universally encouraged as inevitable. Corporations often report that, so far, it does not seem to do much.

Big tech is taking on more debt than ever before to fund its AI aspirations | The Washington Post | 23 Jan, 2026