HubSpot opened Unbound26 with a platform update that changes how marketing and sales teams interact with AI. Instead of picking a separate agent for each task, users can now describe the outcome they want, and the redesigned Breeze Assistant will choose and coordinate the specialised agents required to deliver it.
From a chatbot to an orchestration layer
Breeze Assistant has moved beyond answering questions. A marketer can ask it to fix an underperforming campaign segment, and Breeze can assign content and campaign agents to plan the work, create assets, and personalise nurture sequences. The system can also produce campaign plans, reports and proposals using customer data from the CRM.
Keeping AI context current
That coordination only works if the AI has reliable business context. HubSpot’s new Smart CRM automatically captures and syncs calls, emails and meetings, reducing reliance on manual data entry. A companion feature, Context Home, gives teams a completeness score and shows gaps in the information available to HubSpot’s AI.
The company is sensitive about self-updating data. In July, customer backlash forced HubSpot to reverse a terms-of-service change that would have allowed data from one customer to improve another customer’s records. That history makes transparency and data governance important as Smart CRM expands automatic capture.
What this means for marketers and sellers
Marketing Studio puts the model into practice. It can surface issues such as a weak AEO visibility score, a segment that is lagging, or leads that have not received follow-up. The marketer then tells Breeze what needs fixing, and the assistant coordinates the next step.
On the advertising side, HubSpot says it is the first CRM to integrate with ChatGPT Ads, though Salesforce may contest that depending on how “integrate” is defined. The integration pulls ChatGPT-generated leads into HubSpot for follow-up and attribution. Microsoft Advertising is also being added, with the goal of connecting ad engagement to contacts, pipeline and revenue.
For sales teams, the updated Prospecting Agent monitors more than 40 buying signals, assembles buying groups and drafts personalised outreach. Mobile Notetaker captures conversations, and Deal Progression uses those transcripts to suggest CRM updates and draft follow-ups, with user approval. Revenue Hub can automatically create quotes from deal information already in the CRM.
Why this matters
This is not just another AI feature drop. It signals a shift from AI tools that produce individual outputs to AI systems that own a workflow outcome. For marketing leaders, that means the job description changes: instead of selecting agents, teams will set goals, monitor outputs and audit the data the system uses.
A simple way to prepare:
- Audit CRM data quality before letting agents act on it.
- Run a small pilot on a single campaign or sales sequence, not the entire funnel.
- Define the measurable outcome you want Breeze to improve, such as follow-up coverage or pipeline conversion.
- Keep human approval gates for customer-facing actions while the system builds trust.
The bigger picture
The underlying bet is that usable context, not model choice, will decide which AI agents perform well. HubSpot’s Growth Context combines company, team and customer information so agents can act on a continually updated picture. If the data is stale or incomplete, the orchestration layer will still produce mediocre results. That makes Context Home one of the most practical parts of the update.
Source: MarTech
