MONTEREY, Calif. 鈥� Artificial intelligence is no longer a futuristic buzzword for EMS. It鈥檚 here, reshaping dispatch, billing, documentation and clinical support. But will it revolutionize our systems or replace the humans within them? That provocative question framed one of the liveliest sessions at the California Ambulance Association Annual Conference in Monterey: 鈥淪ix experts 鈥� one weird AI showdown.鈥�
| DOWNLOAD: What paramedics want in 2025
With moderator Danielle Thomas at the helm, six subject matter experts 鈥� Brendan Cameron (ABC), Christian Carrasquillo (Fast Medical AI), Dave O鈥橰ielly (Traumasoft), Nidhish Dhru (Huly), Jonathan Feit (Beyond Lucid Technologies) and Mike Taigman (FirstWatch) took the stage.
The twist? No sales pitches, no product demos; just 2 minutes on the clock once 鈥渂uzzed in鈥� to make their case. The result was a high-energy exchange that was as entertaining as it was thought provoking.
Beyond the buzzer battles and sharp debate, a clear takeaway emerged early: EMS organizations must take a proactive approach to AI. Every agency needs either:
- Internal expertise,
- The ability to hire talent. or
- Access to a trusted consultant who can evaluate solutions for their unique needs
Without those guardrails, the risk is simple: some organizations will be left behind. Jonathan Feit underscored this point by referencing the creation of a federal-level Chief AI Officer role. Risk and compliance managers may not yet have the specialized expertise to evaluate AI adoption, and while the tools may eventually reduce reliance on consultants, there will always be a need for skilled human oversight. As one panelist put it, the job shouldn鈥檛 fall to 鈥渢he best video game player who happens to be a medic.鈥�
Memorable quotes
- 鈥淚f you don鈥檛 change with the time, the time will change you.鈥� 鈥� Nidhish Dhru
- 鈥淭he concern isn鈥檛 AI deciding we鈥檙e competitors. The concern is bad actors using AI to get into our data.鈥� 鈥� Mike Taigman
- 鈥淵ou won鈥檛 lose your job to AI. You鈥檒l lose it to someone who knows how to use AI.鈥� 鈥� Brendan Cameron
- 鈥淧eople are already ingesting things because 鈥楥hatGPT said so.鈥� That鈥檚 job security.鈥�鈥� Jonathan Feit
- 鈥淎I is unlikely to be worse than our current EMS documentation. In fact, it may be better from day one.鈥� 鈥� Mike Taigman
- 鈥淎nything you do manually today 鈥� scanning, attaching, pushing paper 鈥� that job is gone. Not today, maybe tomorrow, but gone.鈥� 鈥� Nidhish Dhru
- 鈥淭ell the truth and nothing but the truth. That鈥檚 how you make AI work for EMS narratives.鈥�鈥� Jonathan Feit
- 鈥淗IPAA violations from AI aren鈥檛 an 鈥榠f.鈥� They鈥檙e a 鈥榳hen鈥� 鈥� and it鈥檚 probably already happened.鈥�鈥� Christian Carrasquillo
- 鈥淲e all want the slam dunk with AI, but in EMS, we haven鈥檛 even learned to dribble, pass or make the layup with the tech we already have.鈥� 鈥� Brendan Cameron
- 鈥淭here are parts of our profession that have a zero margin of error. There is no option. You can鈥檛 kill them again.鈥� 鈥� Jonathan Feit
Fear and the 鈥榡ob鈥� question
The first buzzers made clear the dominant anxiety: will AI take away jobs? Some panelists didn鈥檛 shy away offering that manual, repetitive tasks like billing, scanning and data entry are already vulnerable. Dhru offered a blunt warning:
鈥淎nything you do manually today scanning, attaching, pushing paper 鈥� that job is gone. Not today, maybe tomorrow, but gone.鈥�
Others countered that EMS is chronically understaffed, and AI will simply reallocate resources toward the things humans aren鈥檛 doing well or don鈥檛 have time to do. Cameron pointed out that the real risk isn鈥檛 AI itself but being left behind:
鈥淵ou won鈥檛 lose your job to AI. You鈥檒l lose it to someone who knows how to use AI.鈥�
1 year, 3 years, 5 years
When pressed for timelines, most agreed little will change in the next year. By 3 years, administrative processes and CQI may be significantly augmented. By 5, panelists predicted broader systemic shifts, though no one forecasted full blown 鈥渟uperintelligence鈥� in EMS. Feit took a contrarian angle, noting that AI misuse might actually increase EMS workload:
鈥淧eople are already ingesting things because 鈥楥hatGPT said so.鈥� That鈥檚 job security.鈥�
| MORE: From info overload to instant clarity: Lifeline EMS puts AI to work
Cybersecurity and the 鈥渮ero margin for error鈥�
While job fears dominated early, the debate shifted toward accuracy and risk. Could AI escape human control? Taigman dismissed that notion, saying the more immediate threat is malicious use by cybercriminals. Feit added a sobering reminder:
鈥淭here are parts of our profession that have a zero margin for error. There is no option. You can鈥檛 kill them again.鈥�
Billing mistakes can be fixed; end-of-life wishes or critical medical orders cannot. The panel agreed that accuracy standards must be context specific, and in high-stakes clinical settings, human oversight is non-negotiable.
6 EMS AI applications
The lightning round brought focus to practical uses for AI in EMS:
- Clinical support for differential diagnosis
- Revenue capture through smarter billing
- Data quality checks to highlight gaps
- Process automation to strip out repetitive tasks
- Organizational AI councils to guide adoption responsibly
- Patient-specific insights, surfacing advance directives or autism alerts in real time
Here the conversation turned pragmatic: EMS agencies should not wait for vendors to tell them what鈥檚 possible. Instead, they should train QA/QI staff and designate AI leadership to ensure adoption is safe, compliant and beneficial.
HIPAA, compliance and liability
Compliance loomed large. Carrasquillo sounded the alarm that many providers are already copying patient data into free AI tools without realizing it lingers in external servers:
鈥淗IPAA violations from AI aren鈥檛 an 鈥榠f.鈥� They鈥檙e a 鈥榳hen鈥� and it鈥檚 probably already happened.鈥�
Others warned that the legal system hasn鈥檛 caught up. Liability questions whether errors fall on the vendor, the medic or the agency remain unresolved.
Concerns about AI in EMS: What keeps them up at night
As the final buzzers sounded, the experts revealed their deepest concerns about AI:
- A power shift to big tech companies holding the keys to data and superintelligence
- A dilution of accountability, as people begin to shrug off responsibility 鈥斺�渋t was the AI鈥�
- Unvetted adoption of AI in EMS, with agencies already using models to guide staffing without expertise
- The hype cycle, where EMS leaders chase shiny tools before mastering existing technology
Humor, heat and hard truths
For all its sharp edges, the panel didn鈥檛 lack humor. Feit joked that while AI might not replace EMS roles, it could increase call volume thanks to bad home remedies. Taigman cracked that even flawed AI narratives are 鈥渦nlikely to be worse鈥� than current EMS documentation. And one panelist quipped that if AI ever did 鈥渆nd all suffering,鈥� at least insurance claims would be paid on time.
But beneath the laughs was a serious message: EMS must prepare now. Policies, training, validation and compliance can鈥檛 wait for AI to mature. Technology is already here, and it鈥檚 evolving faster than our rules, laws and comfort zones.
Final takeaways: AI in EMS
Six Experts 鈥� One Weird AI Showdown lived up to its billing: rapid, raw and relevant. With buzzers blaring and 2-minute bursts of insight, the panel delivered a clear takeaway: AI won鈥檛 replace EMS providers overnight, but it will transform the way we work. The question is whether the industry will steer that transformation or be steered by it.
