Authorised fraud
The customer pressed send
APP scams pass authentication because the real customer is acting. Transaction rules see a normal payment to a new payee.
Trust and Identity squad
Chase Vox — Senior AI Conversational Security Agent
AI chatbot security that spots scams mid-conversation
Chase Vox — Senior AI Conversational Security AgentChase Vox is an AI agent that watches your customer chat for scams and social engineering. He spots the signs of a customer being coached by a scammer, checks identity inside the conversation, and hands the chat to the right person with the context attached. Your team decides what happens to the payment.
Chase Vox
Chat session #C-90217
IllustrativeSent to analyst
Scam risk 0.84 · high
Flag explained
“Asks to raise transfer limit, says 'bank security' told them to move funds.”
Job description
What Chase Vox does Job description
Chase Vox is a Senior AI Conversational Security Agent. He sits inside your chat channel, reads conversations for scam and social engineering signs, and passes risky ones to your team with the context attached.
AI AGENT · TRUST AND IDENTITY SQUAD
CHASE VOX
Senior AI Conversational Security Agent
REPORTS TO
Your Head of Fraud
WORKS WITH
Your chat platform, contact centre and payment systems
DEPLOYED
Shadow mode first, then the autonomy you set
KEY RESPONSIBILITIES
01Read customer chats for signs of scams, coaching and impersonation
02Run identity checks inside the conversation before sensitive requests are handled
03Score each chat for scam risk and link it to any pending payment
04Hand risky chats to the right person, with the transcript, signals and payment details attached
05Flag new scam scripts that appear across many conversations
AUTONOMY MODEL
LOWMEDIUMHIGH
Low risk
Chat can continue with a recorded reason, if you allow it
Medium risk
Goes to an analyst by default
High risk
Always goes to an analyst
You set the threshold per rule.
Kill switch: Turn Chase off at any time
Options
How the options compare
| CRITERIA | Agent training and scripts | Keyword filters | Chase Vox |
|---|---|---|---|
| Time to first results | A training cycle | Rule build | Shadow mode on your live chats |
| Who decides | Contact centre agent | Keyword match, then agent | Analyst, inside risk bands you set |
| Reads context | Yes, if the agent spots it | No, single words only | Yes, across the whole conversation |
| Linked to payments | Manual lookup | No | Yes, chat risk attached to the pending payment |
| Hand-off context | Notes typed by the agent | None | Transcript, signals and payment details |
| Where it's weaker | Inconsistent across agents and shifts | Scammers change their wording | Only sees your chat channel, not phone calls or messages elsewhere |
Works best with
AI agents that work best with Chase Vox
Chase's scam signal gets sharper when other agents add what he can't see on his own.
TRUST AND IDENTITY
Nova Sentinel
Lead AI Zero Trust Security Architect
Scores session trust, so Chase knows if the chat comes from a risky device or location.
Meet NovaTRUST AND IDENTITY
Cian Gatekeeper
Lead AI Customer Identity Director
Supplies the customer's linked identity record for in-chat checks.
Meet CianTHREAT
Aiden Flux
Senior AI Fraud Risk Analyst
Scores the payment with Chase's scam signal as an input.
Meet Aiden
Thoughts that matter
Insights on AI agents, financial crime and compliance
What we're learning about AML, fraud and the evidence examiners ask for.
Best AML Software in 2026: An Honest Comparison for Banks and Fintechs
Read More
goAML Registration and STR Filing in the UAE: A Step-by-Step Guide for Reporting Entities
Read More
Transaction Monitoring Software: UAE Acceptance Checklist
Read More
PEP and Sanctions Screening: What the MLRO Owns
Read More
Session Risk Scoring: Continuous Authentication for Banks
Read More
Impossible Travel Detection in Banking: How It Works
Read More
Questions? We Have Answers
-1.svg)
Frequently Asked Questions
Still have questions?
Talk to the people who build the agents. We'll answer per capability, yes or no.
It reads customer chats for scams and social engineering, checks identity before sensitive requests, and hands risky chats to a person with the context attached. Chase Vox links each chat to any pending payment, so your fraud team sees the conversation behind it.
No. Chase scores and flags, and your team decides what happens to the payment. Medium risk goes to an analyst by default and high risk always does. A kill switch turns Chase off without touching your chat platform.
No. Chase sits beside the chatbot and live chat you already run and reads the conversations. Your chatbot keeps answering customers.
He looks for patterns typical of coaching: urgency, secrecy, requests to raise limits or add new payees, and mentions of a 'safe account' or someone claiming to be bank staff. He weighs those against the customer's history and the pending payment.
Chase reads and scores your live chats, but no conversation is interrupted. Your contact centre keeps working as it does today, and you compare his flags with your confirmed scam cases. You decide whether, and where, to switch on any autonomy afterwards.
Chat transcripts with timestamps and customer identifiers, plus pending payment and payee change events. Confirmed scam cases help tune the risk bands.
FluxForce runs as SaaS, on-premise or hybrid, built on Microsoft Azure. We agree data residency and which components run inside your environment during deployment design, before any data moves.