Google launches Gemini Spark, a 24/7 personal AI agent for Workspace users in Australia

What changed: Google introduced Gemini Spark, a personal AI agent that runs continuously on its secure cloud, connecting natively to Gmail, Docs, and Sheets to handle tasks even when a user’s devices are offline. The service is rolling out to Google AI Ultra and Pro subscribers in Australia over the coming weeks, positioning Spark as a shift from reactive chatbot to active work partner.

Why it matters: For operators and knowledge workers, Spark turns routine inbox, document, and spreadsheet chores into background processes, freeing up time for higher‑value work. Founders and IT leaders can treat this as an early testbed for always‑on agents inside existing workflows without custom integration projects.

Try/watch: Start with narrow pilots such as automated email triage, reporting, and document preparation, and closely track how Spark behaves with sensitive data and long‑running tasks.

OpenAI Presence debuts as an enterprise platform for voice and chat agents

What changed: OpenAI launched Presence, an enterprise platform for deploying AI agents that handle complex customer service and internal workflow interactions across both phone calls and text chat. The platform provides infrastructure for organizations to build, deploy, and manage trusted agents rather than stitching together disparate tools.

Why it matters: Presence offers a path to standardize how enterprises roll out agent-based automation, from contact centers to internal help desks, with guardrails baked into the platform. Buyers evaluating AI for customer support can now compare Presence directly with existing contact-center and bot platforms instead of relying on custom integrations around generic models.

Try/watch: Map one high-volume call or chat flow to a Presence pilot, and design clear escalation rules so human agents remain in control while the AI handles routine steps.

Barndoor’s AgentProfile adds identity, cost, and access controls for every AI agent

What changed: Barndoor AI introduced AgentProfile, a capability that gives each AI agent its own profile with a unique record, assigned human manager, and distinct credential sets for tools, models, and data—separate from human user credentials. AgentProfile is available on Barndoor Pro and Enterprise, with per‑agent language model spend and cost attribution through its LLM Gateway, plus a 14‑day free trial of the broader governance and MCP features.

Why it matters: As organizations deploy more autonomous agents, shared logins and opaque spending quickly become operational and compliance risks; AgentProfile targets that by treating agents as first‑class identities with accountable ownership. This helps finance, security, and ops teams see which agents are driving costs and accessing sensitive systems, instead of relying on user‑level reports.

Try/watch: Formalize a policy that every production agent must have an AgentProfile‑style identity, named owner, and scoped credentials, and audit existing automations for shared or orphaned accounts.

Adcore rolls out five autonomous AI agents across its marketing and sales stack

What changed: Adcore announced that five autonomous AI agents are now live across its application suite, exceeding its previous roadmap by delivering more agents ahead of schedule. Four agents—Inbound, Outreach, Deal, and Customer—run on the Proposaly platform to cover the full lead‑to‑revenue cycle, while a Creative Agent in AI Studio generates advertising assets; a sixth Media Buying Agent is slated for launch on the Total Eclipse platform in Q3 2026.

Why it matters: This shows how mid‑market SaaS vendors are packaging full agent teams rather than single features, letting customers automate an entire sales and marketing funnel from first contact to payment collection. For agencies and growth teams, it offers an alternative to assembling separate point tools for lead capture, outbound, proposals, and creative production.

Try/watch: Run a controlled experiment with a segment of leads handled entirely by Adcore’s agents—from inbound response through proposal and onboarding—and compare close rates, response times, and creative quality against human‑only workflows.