Here’s the scene: A council vote is underway. The stream drops. Under SB 707, what happens in the next sixty minutes determines whether that vote is legally defensible.
Starting July 1, 2026, eligible legislative bodies in California must follow a specific, statutory response when a livestream or two-way remote access platform fails mid-meeting. This is a legal requirement with a defined sequence, a minimum wait time, and a formal vote before business can resume. Skip a step, and a court can declare the meeting's actions null and void.
This guide walks through what actually counts as a disruption, the exact protocol your agency must follow when one happens, what your IT team should check during the outage, and how to document it in a way that holds up if it is ever challenged.
A Quick Note on Who This Applies To
We've covered SB 707's core requirements earlier in this series, so here's the short version. The disruption protocol applies to eligible legislative bodies, a category defined by population and revenue thresholds rather than agency type:
- City councils serving 30,000 or more residents
- County boards of supervisors serving 30,000 or more residents
- Any city council located in a county of 600,000 or more residents, regardless of the city's own size
- Special districts with an internet website that meet specific size thresholds: over 200 full-time equivalent employees within a county of 600,000 or more, over 1,000 full-time equivalent employees regardless of location, or over 200 full-time equivalent employees combined with annual revenue exceeding $400 million
If your agency falls into any of these categories, the disruption protocol due by July 1, 2026 is not optional. If you're unsure whether you qualify, that determination is worth confirming with legal counsel now rather than later.
Two Different Kinds of "Disruption"
SB 707 covers two very different situations under the word "disruption," and confusing them leads to the wrong response.
A conduct-based disruption is caused by a person, someone shouting over the proceedings, refusing to yield the floor, or otherwise disrupting the meeting, whether they're in the room or attending remotely. The response is a warning, then selective muting or removal. The meeting continues for everyone else.
A technical service disruption is caused by infrastructure: an internet outage, a dropped Zoom line, a crashed telephone bridge, anything that fails in the systems delivering the meeting to the public. The response is entirely different. The body must stop and recess. All business freezes until the protocol below is complete.
What Actually Counts as a Disruption
Not every technical hiccup triggers this protocol. Based on disruption policies already adopted by California cities implementing SB 707, a qualifying disruption generally has three characteristics:
- Widespread. It affects the public at large, not one resident's home Wi-Fi.
- Ongoing. It persists, rather than a momentary glitch or a single dropped packet.
- Not quickly fixable. Standard adjustments within the flow of the meeting can't resolve it.
One detail trips up a lot of agencies: two-way parity is the baseline requirement. If your video broadcast is running perfectly but your public comment line has crashed, that still counts as a disruption. Residents can watch, but they can't participate, and under SB 707, that isn't good enough. If the public can observe but not comment, the meeting has lost meaningful access and must be paused.
Individual connectivity problems don't trigger the protocol. If one resident's home internet drops, that's not an agency-side disruption as long as your outbound and inbound systems are otherwise functioning.
The Mandatory Response Protocol
Here's what each step actually requires.
1. Announce and recess immediately. The moment the clerk or presiding officer identifies a qualifying disruption, all legislative business stops. The presiding officer announces the outage to the chamber and formally recesses the open session. No votes, no deliberation, nothing agenda-related happens during this recess. The body can still hold an already-scheduled closed session on unrelated items like litigation or labor negotiations.
2. Troubleshoot for at least one hour. The recess lasts a minimum of one hour, or until service is restored, whichever comes first. This is where documented, good-faith effort matters most.
3. Reconvene down one of two paths.
If service is restored within the hour, the presiding officer reconvenes and the meeting continues normally, with full remote access back in place.
If service is not restored after a full hour, the body has exactly two lawful options: adjourn the meeting to a specified future date, or reconvene in person and formally vote to continue without remote access.
The Roll Call Vote You Can't Skip
If your agency chooses to continue the meeting without restored remote access, state law requires more than a quick decision from the chair. The body must formally find, by individual roll call vote, that it made good-faith efforts to restore service and that the public interest in continuing the meeting outweighs the public interest in remote access.
A few procedural details that matter here:
- It can't happen by voice vote or general consensus.
- It can't be placed on a consent calendar.
- Each member's vote must be recorded individually in the minutes.
- If the motion fails to pass, the meeting must adjourn immediately.
This finding is a judgment call the body has to make in real time: is this agenda item urgent enough to justify continuing without full public access, or should it wait? Routine business should generally wait. A time-sensitive bond sale, an emergency declaration, or a statutory deadline is more likely to justify continuing.
What Your IT Team Should Check During the Recess
The one-hour window isn't just a waiting period. It's when your technical staff needs to work through a structured, documented troubleshooting sequence, both to actually fix the problem and to demonstrate the good-faith effort the law requires.
Handing your IT team a vague "the stream is down, please fix it" doesn't produce the kind of record that protects your agency later. A documented checklist does. Broadly, the sequence covers four layers:
Network and connectivity. Confirm the building hasn't lost its outbound internet connection at the router or switch level, and that nothing internal, like a security camera backup or public Wi-Fi traffic, is competing for the same bandwidth as the broadcast.
Platform and software. Check whether the outage is local or part of a broader outage on the platform itself. Zoom, YouTube, Webex, and similar services all publish live status pages. Restart the streaming software. If the connection to the streaming server is being refused, the stream key may need to be regenerated.
Physical hardware. Reboot capture cards and encoders. If audio has dropped but video is fine, check the physical routing on the audio mixer. If a piece of hardware has failed outright, switch to a backup device if your agency has one configured.
Vendor escalation. If none of the above resolves it, contact your platform or IT vendor's support line directly. If a local internet failure is suspected, contact your provider to check the physical line.
Whichever staff member runs this list should keep a timestamped log of every action taken and every vendor contacted. That log becomes part of your official record.
Documentation Is Your Best Legal Defense
Under SB 707, a well-documented disruption is a defensible one. An undocumented disruption is a liability. Your minutes need to capture five specific things:
- The exact time the disruption began and what it was ("6:14 PM, complete loss of inbound audio on the public comment line")
- The procedural actions taken (announcement, recess declared)
- The specific troubleshooting steps attempted, including any vendor or ISP contacts
- The exact time the recess ended and the meeting reconvened, and the total recess duration
- The findings language read into the record and the individual roll call vote results, if applicable
This is the record a court will look at if the meeting's actions are ever challenged. Vague minutes ("technical difficulties were resolved") don't hold up. Specific, timestamped minutes do.
Your SB 707 Disruption Checklist
None of this is a reason to over-engineer your setup or panic every time a Wi-Fi signal blips. There are actionable steps you can follow today to be ready before you need to implement these protocols: a written policy, a tested troubleshooting sequence, and clerk staff who know exactly what to record and when.
A reliable broadcast is still the best way to avoid triggering this protocol in the first place. Multistreaming to more than one destination, so your public comment access doesn't depend on a single platform staying up, is one of the simplest ways to reduce how often you find yourself starting this clock.
Frequently Asked Questions
What counts as a "disruption" under SB 707? A technical service disruption is a widespread, ongoing failure of the two-way remote access required for public participation, one that isn't correctable through routine adjustments during the meeting. This is different from a conduct-based disruption, which is caused by an individual's behavior and handled through warnings, muting, or removal rather than a recess.
How long does an agency have to wait before continuing a meeting after a livestream outage? At least one hour. The legislative body must recess and make a documented, good-faith effort to restore service for a minimum of one hour before it can either reconvene normally (if service is restored) or vote to continue without remote access (if it isn't).
Can a city council continue a meeting if the livestream never comes back? Yes, but only through a specific process. The body must take an individual roll call vote finding that it made good-faith efforts to restore access and that the public interest in continuing outweighs the public interest in remote participation. This finding can't be adopted by voice vote or placed on a consent calendar.
What happens if an agency doesn't follow the SB 707 disruption protocol? Actions taken during a meeting with unaddressed remote access failures can be challenged and declared null and void by a court, requiring the agency to re-notice and re-hold the meeting. A successful challenge can also result in the agency paying the plaintiff's attorney's fees.
Does one resident's internet problem count as a disruption? No. The disruption must be widespread, affecting the public at large rather than an individual's local connectivity, for the statutory protocol to apply.
