FERC’s Software Conference Turns AI Load Growth Into a Grid-Forecasting-and-GETs Story

FERC’s June 5 supplemental notice for its software conference is worth publishing because the useful signal is not simply that another meeting is on the calendar. The stronger signal is that software is being pulled into the core of the grid bottleneck story. When regulators put grid-enhancing technologies, load forecasting, topology optimization, digital twins, and large-load assumptions into one conference agenda, they are telling the market that AI-era power growth can no longer be managed by steel-in-the-ground projects alone.

The official facts are specific. FERC said Commission staff will convene a technical conference on July 7 and 8, 2026 in Docket No. AD10-12-017 to discuss grid-enhancing technologies, load forecasting, and opportunities for increasing market and planning efficiency through improved software. The June 5 supplemental notice says the first day will focus on grid-enhancing technologies and load forecasting, while the second day will cover cutting-edge research topics.

FERC’s June 5 signal is that AI-era power demand is no longer only a steel-and-concrete problem. It is becoming a software, forecasting, and measurement problem too.

What makes this more than a generic software story is how explicitly the agenda ties the issue to the current power context. FERC says the industry is facing rapidly rising power demand, affordability concerns, aging infrastructure, and pressure to integrate new resources quickly. That is already the practical AI-infrastructure environment. The conference framing says software is now one of the tools regulators expect operators to use when they cannot build new transmission, substations, and generation fast enough to match large-load announcements.

Panel one centers on grid-enhancing technologies, but the important read-through is not only the hardware list. FERC is asking how software can identify where latent transmission capacity exists and then manage technologies such as dynamic line ratings, ambient-adjusted ratings, advanced power-flow control devices, and topology optimization once they are deployed. In other words, the conference treats software as the operating layer that may determine whether existing wires can carry more load before capital projects are finished.

Panel two is even more directly tied to AI buildout risk. The notice says rapid load growth, including from data centers, is contributing to load-forecasting challenges nationwide. FERC asks how planners account for uncertain large-load additions, what verifiable commitments should back those forecasts, how operators obtain better data on load size and ramp behavior, and how inaccurate forecasts can force customers to pay for capacity that never turns into real demand. That is not a side question. It is one of the central affordability and planning problems in the current AI boom.

This clears the site’s duplicate block because the thesis is different from the recent FERC governance, large-load, and summer-assessment stories. Those pieces focused on filing rights, cost allocation, reserve margins, or interconnection rules. This story is about the systems layer underneath those fights: whether software, telemetry, and forecast discipline can squeeze more usable capacity out of the existing grid while reducing the odds that utilities overbuild against speculative data-center demand.

For operators, utilities, and developers, the practical implication is that speed to power increasingly depends on data quality and software maturity, not just land, generators, or queue position. A project that cannot provide credible load profiles, ramp expectations, and operational detail is harder to plan around. A utility that cannot integrate better forecasting and grid-enhancing software may leave usable capacity stranded while waiting for slower infrastructure fixes.

For investors and policymakers, the stronger read-through is that “AI power” is becoming partly a measurement problem. The winners may not only be companies that own generation, transmission, or campus sites. They may also include the operators, utilities, and software layers that can turn uncertain large-load demand into bankable planning assumptions without pushing too much speculative cost onto consumers.

The Grid Report view is that this clears the publish bar because it answers a real, high-intent question that is getting more important by the month: why is FERC suddenly treating software and forecasting as part of the large-load story? The answer is that AI demand has moved the bottleneck up one layer. The grid now needs better digital operating systems, not just more physical equipment.

Federal Energy Regulatory Commission, “Supplemental Notice of Technical Conference on Increasing Market and Planning Efficiency through Improved Software,” Docket No. AD10-12-017, published June 5, 2026: https://www.ferc.gov/sites/default/files/2026-06/AD10-12-017%20Tech%20Conf.pdf

Federal Energy Regulatory Commission, “FERC Issues Supplemental Notice for Technical Conference on Increasing Market and Planning Efficiency through Improved Software,” published June 5, 2026: https://www.ferc.gov/news-events/news/ferc-issues-supplemental-notice-technical-conference-increasing-market-and

Federal Energy Regulatory Commission, “2026 Summer Energy Market and Electric Reliability Assessment,” published May 21, 2026: https://www.ferc.gov/sites/default/files/2026-05/26_Summer%20Assessment_0521.pdf