On June 18, 2026, the Federal Energy Regulatory Commission took, in their words, "aggressive, targeted action" on the defining grid question of this decade: how to connect enormous new electricity demand driven by AI data centers while maintaining a reliable, affordable system.

FERC, through a series of individualized "show cause" orders, directed the entities that manage and operate the electric grid to reform their rules and practices to better meet this moment. Among those reforms new transmission services for flexible large loads.

Making AI data centers flexible isn't only what Emerald AI does; it's the key lever to maintaining reliability, driving affordability, and unlocking 100+GW of power.  Which is why FERC's strong message sent on flexibility is so important, and what comes next will be critical.

What FERC did

FERC's orders are not final products; they are tailored directives to our nation's grid operators, requiring each to respond within 60 days. In practice this means there will soon be several parallel proceedings, all running on roughly the same clock, to implement FERC's directive.

The categories FERC identified span issues large loads have highlighted for years, from the need for more efficient processes to cost protection to a requirement for new transmission service products designed for flexible loads.

Why flexibility is critical to FERC's Orders

For most of grid history, a large load was treated as a fixed, immovable block of demand. Under that model, a gigawatt of new load is a gigawatt of new problem.

Flexibility shifts load from a position of static power draw to dynamic grid citizen. An AI factory that can change or shift its consumption in response to grid conditions is not a static block -- it is a resource the grid operator can work with and benefit from. And that single shift in how a load behaves nets a triple benefit: access, affordability, and reliability.

These benefits are not conceptual niceties. FERC's action is premised on the conclusion that flexibility is the difference maker between a data center that waits years for network upgrades and one that can plug in on a timeline measured in months.

FERC's action is the latest (and biggest) domino to fall

FERC's orders are only the most recent example of policymakers seeking to unlock the value flexibility affords. In Santa Clara, the utility Silicon Valley Power is advancing a Flexible Load Interconnection Program in partnership with Emerald AI, which will offer data centers expanded grid access in exchange for verifiable, dispatchable flexibility. Texas is similarly moving in this same direction. The state's grid operator recently finalized a Provisional Controllable Load Resource framework that creates a fast track for flexible data centers to access more power from the grid ahead of traditional interconnection timelines, in exchange for being controllable when the system needs it. And in the SPP footprint, the Conditional High Impact Large Load Service (CHILLS) path, part of the FERC-approved framework for high-impact large loads, offers conditional, curtailable grid access on an accelerated study timeline. Different region, different mechanism, same underlying principle: a load willing to flex earns a faster, more efficient path onto the system.

What happens next

Grid operators have 60 days to respond to FERC's orders. As they do, interested groups will have an opportunity to shape the details: what counts as a flexible load, what operational performance is required, how the value of flexibility is recognized, and what protections apply to loads that commit to it.

FERC's action makes clear that flexibility is a critical lever for enabling load growth, reliability, and affordability all at once. Emerald is already proving the model works in the field, and will be helping to now drive the opportunity FERC has created into reality on the ground.

‍