OpenAI is intensifying its dual focus on artificial intelligence infrastructure and enterprise software. The creator of ChatGPT has announced plans to construct a massive data center in Georgia, USA, with a total investment exceeding $30 billion and a planned power capacity of 3.2 gigawatts. Concurrently, it has launched an enterprise AI agent platform named Presence, marking a significant expansion into the corporate software market.

On July 22, reports indicated that OpenAI has signed a power supply agreement with Georgia Power, securing 3.2 gigawatts of capacity for the Georgia-based data center. The first few hundred megawatts are expected to come online in 2028, with the entire project slated for completion by 2032. Sachin Katti, Vice President of Compute Strategy, stated that if built to its full scale, the total investment would surpass $30 billion, and the company is currently seeking development partners.

In a parallel move, OpenAI has officially introduced the Presence enterprise AI agent platform. This platform deeply integrates AI models with a company's internal data, policies, and business processes, offering automated solutions for scenarios like customer service, sales, and IT support. This represents a shift beyond simply selling model access.

As competition in foundational models heats up and prices continue to fall, OpenAI is simultaneously expanding its computing power investments and extending its reach into enterprise software services. This dual-pronged strategy of building both infrastructure and applications aims to forge a new competitive edge.

Building a 3.2GW Data Center with Over $30 Billion

The Georgia project stands to be one of OpenAI's largest AI infrastructure investments to date.

Under the agreement, OpenAI will secure 3.2 gigawatts of power capacity. The first few hundred megawatts are expected to be phased into use starting in 2028, with the full construction period extending to 2032. This 3.2-gigawatt scale is equivalent to the electricity demand of approximately 2.4 million U.S. households.

Katti noted that if the project proceeds at full scale, the total investment will exceed $30 billion. Currently, OpenAI has committed $20 billion to Effingham County, Georgia, in exchange for a package of local tax and incentive policies pending approval. The company is also in talks with potential partners regarding financing and joint development and may operate some facilities independently.

Reports suggest that as AI computing power demand continues to surge, OpenAI has revised its future cloud infrastructure spending forecast upward to $750 billion, indicating a sustained increase in long-term capital investment.

From Participant to Leader in "Stargate"

The Georgia project also reflects a strategic repositioning by OpenAI regarding its Stargate initiative.

Stargate was initially launched jointly by OpenAI, Oracle, and SoftBank. However, as the project advanced, collaboration models across campuses have diversified, and OpenAI has begun to take on more planning and design responsibilities.

Katti explained that after building two to three generations of AI infrastructure, the company has accumulated significant design expertise. This allows for further optimization of data center architecture to reduce costs and accelerate construction speed, leading to the decision to lead the overall design in more projects.

OpenAI also plans to gradually open its AI data center design standards, including solutions for data centers, servers, and network architecture, aiming to build an open ecosystem similar to Meta's Open Compute Project.

Simultaneously, the company is adjusting its global investment portfolio. It has reportedly exited some data center projects in the UK and Norway and paused expansion at its flagship Stargate campus in Texas, concentrating resources on larger, more strategically significant infrastructure projects.

Proactive Community Engagement

The rapid expansion of large-scale data centers has also sparked increasing debate concerning power, water usage, and environmental impact.

In response, OpenAI has proactively introduced several supporting measures. The company stated it will cover the construction costs of associated power transmission and distribution facilities, preventing these expenses from being passed on to local residents. The data center will actively reduce its electricity consumption during periods of high grid load and will employ a closed-loop recirculating water cooling system to minimize water usage.

Furthermore, the company has committed to paying hundreds of millions of dollars in taxes to state and local governments, establishing a community fund, and providing local students with usage credits for OpenAI products.

Chris Lehane, OpenAI's Chief Global Affairs Officer, said the company will begin ongoing communication with local residents this week and will commission independent third-party oversight audits to ensure the project adequately addresses community concerns.

Presence Targets the Enterprise Software Market

While continuously expanding its infrastructure, OpenAI is also accelerating its enterprise-grade product deployment.

The newly launched Presence platform goes beyond simply providing model APIs. It deeply integrates AI models with a company's internal data, policies, existing software, and business processes, enabling businesses to deploy AI agents capable of autonomously completing tasks.

The platform primarily targets scenarios such as customer service, sales, and IT support, capable of automating specific business processes like billing inquiries, insurance claims processing, and employee IT service requests.

Compared to merely selling model access, this type of enterprise software product offers higher customer retention and can help OpenAI establish a more stable long-term revenue stream. Currently, Presence operates on a customized deployment model, requiring businesses to work with OpenAI engineers or designated partners for implementation. OpenAI revealed it has already piloted the platform for its own English-language customer service hotline.

As competitors like Anthropic, Google, Meta, xAI, Mistral, and DeepSeek continue to enhance model capabilities and drive down prices, foundational models are gradually becoming commoditized. For OpenAI, deeply embedding AI into enterprise workflows to provide complete solutions, rather than just selling model access, is becoming a crucial competitive focus for its next phase.