Group I-2 Is Risk Category IV: the 'Condition 2' Table Fix

The two words that disappeared

Here's the thing: the January 2026 errata to the 2025 California Building Code did something tiny to the Risk Category table. It deleted two words. In Tables 1604.5 and 1604A.5, the errata struck the words "Condition 2" from the Group I-2 line in the Risk Category IV description; that line now reads simply "Group I-2 occupancies."

Two words. But those two words could be read to argue a hospital wasn't Risk Category IV — "my building isn't a 'Condition 2,' so this line doesn't catch it." The errata closes that door, because the words never described anything real in California to begin with.

Why 'Condition 2' never belonged in an I-2 line

California does not use the model IBC's I-2 "Condition" subclassification. Open Chapter 3 and the proof is right there.

Section 308.3 defines Institutional Group I-2 — medical care on a 24-hour basis — and lists what falls in it: foster care facilities, detoxification facilities, hospitals, nursing homes, and psychiatric hospitals. No Condition 1. No Condition 2. The group simply isn't subdivided.

The Conditions live one section over. Section 308.4 classifies Institutional Group I-3 — detention and correctional buildings under restraint or security — and that group is split into Condition 1 through Condition 9 (308.4.1-308.4.9), describing how locked the egress is. California even adds Conditions 6 through 8 as state amendments.

So "Group I-2, Condition 2" was a phrase with no referent. There's no Condition 1 I-2 to contrast it against. The term rode in from the model code and survived the first printing of the 2025 edition until the errata caught it.

Verified against the live code text: CBC Table 1604.5, Risk Category IV, now reads "Group I-2 occupancies." And §308.3 lists hospitals, nursing homes, psychiatric hospitals, detox and foster-care facilities under Group I-2 with no Conditions, while §308.4 reserves Condition 1 through 9 for Group I-3 detention. (2025 CBC, Part 2, with January 2026 errata.)

What the corrected line drives

This is where the structural stakes show up. Risk Category isn't a label — it's an input to the loads. Section 1604.5 says the risk category from Table 1604.5 is used in lieu of ASCE 7, Table 1.5-1. From there it feeds the seismic importance factor, Ie, the Seismic Design Category, and the wind design. The earthquake-design-data a designer must show on the drawings (CBC §1603A.1.5) opens with exactly two lines: Risk category and Seismic importance factor, Ie. One drives the other.

With the table now reading "Group I-2 occupancies," the whole class lands in Risk Category IV — the same essential-facility tier as fire stations, emergency operations centers, and water-treatment plants. A Group I-2 building gets designed to that standard, and the deleted words can no longer be read as a loophole to drop it a tier.

The OSHPD wrinkle — confirm it with your AHJ

One honest caveat, because the table carries OSHPD annotations. In the base Table 1604.5, the "Group I-2 occupancies" line is flagged "[OSHPD 2 & 5] Not adopted by OSHPD," and those facilities — skilled nursing and intermediate care (OSHPD 2) and acute psychiatric hospitals (OSHPD 5) — appear instead in the Risk Category III rows where they have 50 or more care recipients. Meanwhile the HCAI-enforced Table 1604A.5 lists General Acute-care Hospital Buildings (OSHPD 1 & 4) explicitly in Risk Category IV, right alongside "Group I-2 occupancies."

The practical read: a general acute-care hospital building is Risk Category IV. For a skilled-nursing or acute-psychiatric project, which table and which row governs turns on the enforcing agency and the care-recipient count — so confirm the controlling table with your jurisdiction or HCAI before you set the importance factor.

The same January 2026 errata also pulled those stray Condition tags out of the I-2 life-safety sections — smoke control, audible alarms, exit illumination — where California already governs by smoke compartment. That's a companion story. This one is structural: the loophole the deletion closes is a Risk Category one.

What CrossBeam does with it

CrossBeam reads a Group I-2 project against the Risk Category the California table actually assigns — and against the OSHPD note that decides which table governs — so the seismic importance factor and Seismic Design Category trace to the live code text, not to a model-code term California never carried. When an errata quietly deletes two words, the analysis follows the corrected code, not the leftover.