Howdy! I've split this fortnight's AI news roundup into three sections so you can skim what matters and act on it. Tech covers the big platform moves that change the baseline. Tools highlights features and products you could actually switch on inside the apps you already use. Trends distils what this means for Australian SMEs — policy signals, risks, and the budget case for 2026.
TECH
1) OpenAI goes multi-cloud — a 7-year, US$38B deal with AWS
OpenAI signed a seven-year, US$38 billion cloud agreement with Amazon Web Services. Practically, that means far more training and serving capacity for the next generation of models and agentic systems. It also breaks the perception that OpenAI is tied to a single cloud (Microsoft's Azure). For SMEs, the signal is continuity: the underlying models you touch through SaaS and office suites are likely to keep improving quarter-on-quarter through 2026. That argues for a steady adoption plan (governance + training + workflow changes), not a one-off experiment.
2) Google's AI focus is "in your existing tools," not sidecars
Google's October/November drops emphasised Gemini Enterprise and Workspace integrations. AI inside Gmail/Docs/Sheets/Meet with admin-grade controls, rather than separate apps staff must learn. If you're a Google user, expect more switches to appear in the admin console and a gradual uplift in "generate/explain/summarise" capabilities across documents, meetings, and knowledge retrieval. Planning task: align your acceptable-use policy with what Gemini can now do, and decide which teams get it first.
3) "AI on your data" keeps maturing (SAP, Snowflake, Databricks)
At SAP TechEd, SAP announced zero-copy integrations and a push to make business data directly usable for AI. Snowflake showcased document-scale analytics that goes beyond classic RAG by querying thousands of documents as structured data. Databricks added governance and evaluation tools for AI agents. The shared theme: keep your data where it lives, minimise copying, and govern access centrally. If you're already on one of these stacks via a partner, ask specifically about document analytics, agent governance, and zero-copy patterns.
TOOLS
4) Dynamics 365 Business Central: more native AI in finance/ops tools
Microsoft's current release wave for Business Central includes planned features that embed AI into everyday tasks — generate or explain entries, summarise documents, and add assistant-style prompts inside the finance/ops backbone. The advantage for SMEs is adoption friction: staff use AI where they already work, rather than jumping to another app. Ask your partner which features are GA in your tenant and what training templates they provide.
5) The SME stack is converging: AI assistants in service desks and finance
Two signs worth noting. First, Zendesk is rolling out "AI agents" that can resolve a growing share of support requests with policy/knowledge controls for admins. Second, Xero has been expanding Just Ask Xero (JAX) to behave more like a finance "super-agent." The pattern is clear: core line-of-business tools are shipping their own assistants. Your move is to standardise on a few trusted, admin-controlled assistants and train teams on those — not on dozens of one-off tools.
TRENDS (Australia)
6) Canberra's line on Governance is pragmatic: have a policy, prove control
The federal Guidance for AI Adoption emphasises practical steps for organisations: set a one-page policy, define data rules, document human-review points, and revisit quarterly. In parallel, whole-of-government policy work reiterates a risk-based approach. For SMEs, this validates a lightweight governance model that's easy to teach and audit. If you don't yet have a one-pager and R/A/G data rules, that is the fastest-ROI step you can take before switching on more features.
7) The CFO case for AI investment is strengthening: up to A$142B a year by 2030
A new OpenAI-funded economic blueprint estimates AI could add up to A$142 billion per year to Australia's GDP by 2030 if adoption accelerates — with the bulk coming from businesses applying AI to operations, not from building models. Other local analyses (Microsoft/TCA, and the Productivity Commission) put the upside in a similar ballpark. The takeaway for budgeting: position AI line items against productivity uplifts and revenue enablement, not "innovation spend," and track realised gains in cycle time, quality, and working capital.
Want help turning this into a 90-day plan?
If you're on Google or Microsoft and saw yourself in any of the items above, we can map where to start: which assistants to enable, which workflows to standardise, and what policy/training to put in place so adoption sticks. We'll turn that into a simple 90-day plan your team can execute and we'll stay on call to adjust as features roll out.
Cheers,
Patrick
---
Relevant Insights
- Four Microsoft Copilot Features Australian SMEs Should Actually Be Using — High-value Copilot features that solve real problems
- AI Strategy Essentials for SMEs — Develop a clear, actionable AI strategy tailored to your business goals
- Executive Guide to AI Governance for SMEs — A simple approach to AI governance that enables responsible use
---
About the Author
Patrick is the co-founder of The AI Guides and helps Australian leadership teams turn AI from buzzword to business results, drawing on a decade in strategy and digital transformation.
About The AI Guides
The AI Guides is a Sydney-based AI advisory partner for Australian SMEs, focused on strategy, training, implementation support and lightweight governance.