We ship
So you stay ahead.

See how we are raising the bar for AI agents. Stay up to date with our releases, built to make your AI agent faster, more accurate, and more capable than ever.

May 13, 2026

NEW

Dynamic Safeguards: Context-aware responses when safeguards trigger

Replace the fixed safeguard fallback with an actionbook that generates the response in real time, branching by trigger type and using the full conversation context.

May 7, 2026

Zero-Touch Improvement: Autonomous AI agent optimization with human oversight built in

Your AI researches its own failures from real conversations, proposes precise fixes, tests them, and deploys at whatever level of autonomy your team has approved.

May 7, 2026

Agent Steward: AI that owns complex customer issues end to end

The first AI agent that takes full ownership of complex customer issues—coordinating across systems, channels, and time until the case is closed.

May 7, 2026

A/B Testing: Know whether your AI agent changes actually worked

State a hypothesis, run two versions of your agent against real traffic, and get a statistical reliability score—not just "no regressions."

Apr 28, 2026

Openbook: Visual flow navigation for your Actionbooks

See your entire actionbook as a visual flow then jump to any section in one click.

Apr 22, 2026

AI Agent Urgent Notice: Push real-time notices without a deployment

Support teams can now publish real‑time Urgent Notices directly from the AI agent dashboard, alerting customers instantly without waiting for a deployment.

Apr 13, 2026

Actionbook Editor: A reliable WYSIWYG environment for authoring your AI’s rules

A single editing surface for your AI's rules, with a live Viewer panel so you always see what your AI will see before you save.

Apr 9, 2026

Test Upgrade: Build your AI agent test suite from real conversations

Build your test suite from real conversations, run it anywhere, and get failure analysis that tells you where to look.

Mar 26, 2026

Thinking Messages: Real-time status updates while your AI agent works

Show users exactly what your AI is doing with a specific status message instead of a generic typing indicator.

Mar 25, 2026

Human Initiated Proactive Email in Desk: Reach customers first, without leaving your support queue

Stay ahead of customer issues with proactive emails—send updates on delays, downtime, or billing changes directly from your Desk dashboard.

Mar 10, 2026

Gradual Rollout: Test with real traffic, commit when you're confident

Discover how Gradual Rollout lets teams validate actionbook and model changes against live traffic before committing to a full deployment.

Mar 3, 2026

Sound Notification: When a new message arrives, users hear it

Learn how Sound Notification alerts users the moment an AI agent replies, using audio or mobile push so conversations stay active even when switching tabs.

Mar 1, 2026

Low CSAT Analysis: Filter bad conversations by failure type

Low CSAT Analysis brings AI-powered failure categorization into the Conversations tab, simplifying the process of finding and filtering conversations.

Jan 26, 2026

Desk Glossary: Translation without rules is just expensive guessing

Brand names, job titles, compliance terms, regional variations—every industry has language that can't be left to guesswork. Now you control exactly how it translates.

Jan 7, 2026

Staging Environment: A new dedicated testing layer between dev and production

Staging Environment gives your team a production-mirror space to validate AI agent changes with real settings before deployment, eliminating production surprises.

Dec 17, 2025

Selective Deployment 2.0: You had control, now you have visibility

Selective deployment gets smarter: Eliminate deployment anxiety with full visibility into every change before it reaches production.

Dec 2, 2025

Activity Log 2.0: From black box to open book

Reveal the complete reasoning behind every agent response—which actionbook, tools, and personalized instruction shaped each answer—so you can verify accuracy and improve with confidence.

Nov 26, 2025

Conversation Search: Find any conversation, instantly

Search thousands of customer conversations to quickly surface the specific interactions you need to investigate, troubleshoot, or analyze.

Nov 18, 2025

Omnichannel continuity: Turn every pause into progress

Omnipresent AI that follows customers across every channel, picking up exactly where they left off to create frictionless, consistent experiences.

Nov 18, 2025

For You Conversations (FYC)

Deliver personalized, adaptive AI interactions tailored to each customer’s interests and behavior.

Nov 18, 2025

Agent Memory Platform (AMP)

Your AI can finally remember every customer and know exactly how to help.

Oct 31, 2025

Collect instant user feedback on every AI agent message

Gather instant thumbs-up/down feedback on every AI agent response to uncover actionable insights and continuously improve accuracy.

Oct 16, 2025

Embed interactive forms in agent responses

Empower your AI agents to collect information seamlessly with interactive forms that make every customer conversation faster and more efficient.

Sep 30, 2025

Customizable post-chat CSAT survey

Not every business measures satisfaction the same way. Take control of customer feedback with customizable post-chat CSAT surveys that are tailored to your brand, your rules, your channels.

Sep 29, 2025

Sendbird AI agent is ISO 42001 certified

Sendbird is now ISO 42001 certified and committed to building responsible AI agents with industry-leading governance, security, and risk controls.

Sep 23, 2025

Humans as AI evaluators

AI can’t define good, people do. Evaluate conversations on your criteria. Teach AI what good looks like.

Sep 12, 2025

Multiple active conversations per customer

Faster answers, cleaner reporting, happier customers. With multi-conversation support in the AI agent messenger, your customers can start and maintain multiple ongoing conversations simultaneously.

Sep 9, 2025

Desk AI Copilot

Every ticket is a race against time. Desk AI copilot helps you stay ahead.

Aug 26, 2025

Voice AI agents: Faster, smarter, and built with empathy

Entering the new era of Voice AI — designed for the chaos of real conversations.

Aug 13, 2025

Citations: AI answers you can trust and explore

Your AI agent now delivers more than just answers, it creates pathways to deeper knowledge.

Aug 8, 2025

GPT-5 now live in Sendbird AI agent

We’re excited to bring GPT‑5 to production inside the Sendbird AI agent platform, powering deeper reasoning, better context retention, and more reliable automation where it matters most.

Jul 30, 2025

Trust OS: The foundation for responsible AI agents

Unlock full visibility, control, and oversight so your agents stay safe, compliant, and trustworthy.

Jul 28, 2025

Automatic message translation in Sendbird Desk

Break language barriers in real-time with automatic message translation in Sendbird Desk. Deliver seamless multilingual support from AI to human agents—without switching tools.

Jul 23, 2025

Outcome-based testing: The end of shallow AI validation

Most AI testing stops at “Did it sound right?” But customers care about results. Outcome-Based Testing goes deeper, it validates whether your AI understood, acted, and actually solved the problem.

Jul 16, 2025

Data export API: Turn AI agent conversations into actionable business intelligence

Enterprise teams need more than chat logs, they need data that drives decisions. Export AI agent performance metrics, conversation details, and CSAT scores directly to your business intelligence tools.

Jul 1, 2025

Selective deployment & rollback: Built for control in complex teams

You build carefully. Now you can deploy the same way.

Jun 30, 2025

Flagged message monitoring

See when users try to break your AI, and how well your safeguards hold up.

Jun 17, 2025

One-click convert to knowledge: Teach your AI agent from real conversations

Stop losing your best AI training data to chat history. One click turns any conversation into reusable knowledge that makes your AI agent smarter instantly.

May 22, 2025

AI hallucination detection & alerts

AI hallucinations aren't a matter of if, they're a matter of when. The only question is whether you’ll catch them first, or hear about them from frustrated customers.

May 13, 2025

Sendbird localized AI agent built for real-world scale

Localization isn’t just a language problem, it’s a retrieval problem. Sendbird’s localized AI agent aligns every answer and action with the region, the language, and the reality on the ground.

May 8, 2025

Deployment and production mode: Stay in control at scale

AI is never static. Give your team the speed to build, the structure to stay safe, and the precision to scale cleanly.

May 1, 2025

AI agent scorecards: Bring oversight and accountability to AI

Good AI isn’t self-correcting. Good AI is managed. Give your AI agents the same operational discipline you expect from human support teams.

Apr 30, 2025

Activity trails: Full visibility into how your AI agent acts

Trust isn’t just about the outcome, it’s about knowing exactly how your AI agent got there. With Activity Trails, there’s no black box.

Apr 28, 2025

End-to-End AI agent testing — Build trust before you go live

Sendbird’s E2E Test simulates real conversations from start to finish, so you can launch with total confidence in every message, every action, every customer.

Apr 2, 2025

Sendbird’s unified support: AI agent & Desk

Cut the clutter. Bring messaging, ticketing, and AI agent together in one lean platform to simplify your support stack.

Mar 28, 2025

Proactive Chat: AI that reaches out first

Proactive Chat gives the AI agent the power to move first, reaching out before frustration builds, before churn begins, before a ticket even exists.

Mar 25, 2025

Sendbird AI agent on WhatsApp

We’re bringing the Sendbird AI agent to the world’s most popular messaging app. You can now deploy your AI agent on WhatsApp to deliver fully automated, human-like customer support, without customers ever leaving the app.

Mar 20, 2025

Salesforce supported as a knowledge source

Empower your Sendbird AI agents with real-time answers from your Salesforce knowledge source—no manual updates, just faster, accurate customer support.

Mar 20, 2025

Sprinklr integration for AI to human handoff and knowledge

Support gets smarter and more seamless with the Sendbird AI agent + Sprinklr integration. Now you can escalate conversations to human agents directly in Sprinklr, while also training your AI using your existing Sprinklr knowledge base.

Mar 18, 2025

Multi-language AI agent that not just translates

Create localized AI agents that speak the right language with cultural awareness, delivering a more natural, human-like customer experience.

Mar 18, 2025

Easily manage and track versions of your AI Agent Action Book to refine AI workflows, improve resolution rates, and improve performance over time.

Mar 14, 2025

Sendbird AI agent team — specialized AI agents for every role

Build a team of specialized AI agents with the Sendbird AI agent platform for more efficient customer support.

Mar 13, 2025

Sendbird AI agent now available on email

With Sendbird AI agent for email, your team can deliver instant, intelligent, and fully automated support—directly through your existing email channels.

Mar 11, 2025

Freshworks integration for AI to human handoff

Connect Sendbird’s AI agents with Freshworks for seamless AI-to-human handoff when complex issues require expert human support.

Mar 3, 2025

Sendbird AI agent on SMS

Sendbird’s AI agents are now available on SMS to deliver proactive, AI customer service for faster, more engaging support.

Feb 4, 2025

Confluence now supported as a knowledge source

Connect Sendbird AI agents with Confluence to deliver accurate, expert answers using your company’s data and knowledge base.