April brought a steady mix of AI evolution, operational improvements, and a few changes that need proper attention: especially if you’re running bots or investing in automation.
Here’s a clear breakdown of what’s changed, what matters, and what to act on.
The big shift: AI agents are becoming the default layer
The most important development this month is how Zendesk is pushing AI agents deeper into core workflows, especially email and ticket creation.
Agentic AI for email is now generally available
Advanced email AI agents can now:
- Understand full email context (including messy threads)
- Answer multiple questions in one go
- Execute actions (like cancellations) via integrations
- Reduce back-and-forth by gathering all required info upfront
What this means technically:
- Less reliance on flow builders and triggers
- More dependency on knowledge quality and system integrations
- A shift from “automation logic” → “AI reasoning + procedures”
What to watch:
- Your knowledge base becomes critical infrastructure
- Poor content = poor AI outcomes
- Testing edge cases (multi-intent emails, vague queries) is now essential
⚠️ Mandatory change: AI agent tickets for bot conversations
From May 4, all bot interactions (including third-party bots) will automatically create AI agent tickets.
Why this matters:
- Every interaction is now stored, structured, and visible
- Better compliance (GDPR, audit trails)
- Full reporting across bots + humans
The risk: duplicate tickets
If you currently:
- Create tickets via bot platforms
- Use webhooks or custom integrations
You must update your setup.
Required actions:
- Audit any “create ticket” logic in your bots
- Remove or disable duplicate ticket creation
- Test in sandbox before rollout
- Train agents on the new ticket format
Key takeaway:
Zendesk is centralising conversation tracking. Workarounds and parallel systems will start to break.
Messaging and agent experience improvements
Audio playback directly in Agent Workspace
Agents can now play audio messages without downloading files.
Impact:
- Faster handling of voice-based conversations
- Better accessibility support
- Cleaner workflow (no external tools)
No setup required. It’s live automatically.
Malware scanning for messaging attachments
All files shared via messaging are now automatically scanned.
What changes:
- Suspicious files are blocked until reviewed
- Admins/roles can approve flagged content
Why it matters:
- Stronger security baseline without extra tooling
- Particularly relevant for social and web messaging channels
User suspension now supports messaging-only restrictions
You can now suspend users only from messaging channels, instead of globally.
Use cases:
- Handling spam or abuse in chat without affecting email support
- More granular control over customer access
Search, knowledge, and content improvements
Customisable filters in advanced search
Search in Support is now more flexible:
- Add/remove filters via a side panel
- Tailor views to specific workflows
Practical benefit:
- Faster debugging, reporting, and ticket analysis
- Less dependency on fixed views
Notion as a knowledge source
You can now use Notion as an external content source.
Why this matters:
- Knowledge no longer has to live only inside Zendesk
- Easier to reuse product docs, internal guides, or ops content
But:
- Content structure and governance become more important
- Sync quality directly impacts AI answers
Updated article viewer in Web Widget
A refreshed article experience improves how users consume help content inside the widget.
Impact:
- Better self-service experience
- Potential reduction in ticket volume if content is strong
Summary: what actually matters this month
1. AI is moving from optional to foundational
- Email AI is now capable of end-to-end resolution
- Bot interactions are fully integrated into ticketing
👉 This is not just a feature release, it’s a platform direction.
2. You need to review your bot architecture (urgent)
- Duplicate ticket risk is real
- Legacy integrations will conflict with new defaults
👉 If you use third-party bots, this should already be on your roadmap.
3. Knowledge quality is becoming a technical dependency
- AI performance relies on structured, accurate content
- External sources (like Notion) increase flexibility—but also risk
4. Small UX improvements are adding up
- Audio playback
- Better search
- Improved widgets
👉 These won’t transform your setup alone, but they improve daily efficiency.
Key takeaways
- Act now on AI agent ticket changes before May 4
- Audit automation logic—especially anything creating tickets
- Invest in knowledge management, not just automation
- Expect AI to replace more deterministic workflows over time