AI drafts and triages tickets fast enough to cut first response time in half, but it still mishandles or wrongly closes at least 1 in 10 refund and account access requests without flagging the miss.
Intercom Fin, Zendesk AI, Claude, and ChatGPT were tested on ticket triage, reply drafting, and deflection tasks. AI for customer support scored 3.4 in August 2026.
- ✓Ticket triage and routing cut first-response time from 40 minutes to under 10
- ✓Reply drafting from tickets plus knowledge base saves 60-70% per reply
- ✓Deflection on known issues resolves resets and status checks without a human
- ✓Auto-tagging closed tickets into themes frees up a day per week
- ×Refund and account security requests bypass fraud review on pattern matches
- ×No tool tracks multi-ticket context across one customer over time
- ×Escalation replies ignore account value and churn risk by default
Scorecard
Handles routine tickets well but mishandles roughly 1 in 10 edge cases.
Integrates into existing ticketing tools and cuts first-response time significantly.
Treats each ticket in isolation, missing multi-ticket customer history.
Agents edit AI drafts instead of writing from scratch, saving 60-70% per reply.
Does not flag when pattern-matching overrides judgment on refunds or escalations.
The unsolved problem is per-customer memory across tickets, agents, and tools. A solution would maintain a living case file per customer (past tickets, resolutions, refund history, account value) attached to every new ticket, and flag when a new ticket matches a prior unresolved issue.