Claude AI is an AI assistant built by the American company Anthropic. It handles text, documents, code and image analysis in one interface, and comes as a family of models: Fable 5 (frontier), Opus 4.8 (powerful all-rounder), Sonnet 5 (balanced) and Haiku 4.5 (fastest). Australia sits in the global top four for Claude usage per capita.

Claude is owned by Anthropic, an American AI company founded in 2021 by siblings Dario and Daniela Amodei together with several former OpenAI staff. Anthropic has major investors including Google and Amazon, but remains an independent company focused on AI safety. In 2026 it opened an office in Sydney.

Anthropic is owned by its founders and employees together with a number of large investors. The company was founded in 2021 by siblings Dario and Daniela Amodei plus a group of former OpenAI staff, and is headquartered in San Francisco. Google and Amazon are the biggest outside investors (Amazon has invested an estimated four billion US dollars in AWS resources and cash), but neither of them owns Anthropic. They're investors in an independent American company focused on AI safety.

How do I get started with Claude AI?

Three paths. Create your own account at claude.ai (free, or Pro at US$20 a month). Buy Team licences directly from Anthropic from US$20 per user per month with annual billing (US$25 monthly, minimum 2 seats). Or take a managed option like satori-claude where we handle licences, support and training with one invoice in AUD.

Yes, there's a free version at claude.ai with limited usage. For daily business use it rarely stretches far enough. Pro costs US$20 a month (roughly A$30), Team Standard from US$20 per user per month with annual billing (roughly A$30) and Enterprise starts at US$20 per seat self-serve plus usage-based costs, or a quote for sales-assisted (minimum 50 seats).

Does Anthropic have an office in Australia?

Yes. Anthropic opened its Sydney office in 2026 and has signed a memorandum of understanding with the Australian government. Australia sits in the global top four for Claude usage per capita, so the local investment isn't surprising.

What is Claude AI? The complete Australian guide | satori.

Q: Is Claude better than ChatGPT?

It depends on the work. Claude leads on coding, legal work, financial analysis and long documents thanks to a model family built for long context. ChatGPT is stronger on image generation and voice. For text-heavy business work Claude is usually the better fit, and plenty of companies simply run both.

Q: Is Claude safe for company data?

On paid business plans Anthropic doesn't train on your data by default, a data processing agreement is part of the commercial terms, and the company holds SOC 2 Type II, ISO 27001:2022 and ISO/IEC 42001:2023. For Australian Privacy Act purposes, Claude is also available through AWS Bedrock in Sydney (ap-southeast-2) if data residency matters to you.

Something odd is happening down here. Australia, a country of 27 million people, sits in the global top four for Claude usage per capita.¹ Anthropic noticed. In 2026 they opened an office in Sydney and signed a memorandum of understanding with the Australian government along the way.²

But the question we get most often when we sit down with IT managers is still the simplest one. What is Claude AI? Who builds it? How does it differ from ChatGPT, the name everyone already knows? And why has this particular tool become the one Australian businesses keep landing on?

This guide collects the answers. Not as a technical deep-dive. As answers you can pass on to the HR manager, the controller or the sales rep at the next desk.

And if you're already running Claude? Good. This guide helps you explain to someone else why.

What is Claude AI? The short answer

Claude AI is an AI assistant built by the American company Anthropic.³ It's not a search engine. And not a chatbot in the narrow sense. It's a family of large language models plus a growing ecosystem of tools that together handle text, documents, code, image analysis and, these days, design work too.

The model family is the brain. Fable 5 is the frontier model. Opus 4.8 is the powerful all-rounder. Sonnet 5 is the balance of smart and fast. Haiku 4.5 is the quickest and cheapest. More on all of them below.

What separates Claude from other AI assistants is something Anthropic calls Constitutional AI. A method for training the model to follow an explicit set of principles, including parts of the UN Declaration of Human Rights. In 2026 the "constitution" runs to 23,000 words. In 2023 it was 2,700.⁴

In practice, that means Claude says no to harmful requests without seizing up over every grey area. And that it's genuinely good at reasoning, planning and holding on to complex tasks without losing the thread.

For Australian businesses, the first answer is this. Claude is an AI built with safety and reasoning at the centre. Not speed or viral demos.

Who owns Claude AI?

Claude is owned by Anthropic. Full stop. But the story behind it is more interesting than that.

Anthropic was founded in 2021 by siblings Dario and Daniela Amodei together with a group of former OpenAI staff. They were the team behind GPT-3 at OpenAI. And they walked away to build something they thought mattered more. AI with safety at the centre from day one.

Today the company has serious backers. Google has invested billions of dollars. Amazon has invested even more, an estimated four billion US dollars in AWS resources plus cash.⁵ Neither of them owns Anthropic. They're investors. Anthropic remains an independent American company headquartered in San Francisco.

And since 2026, one with a Sydney office.² Local presence means local staff, local partnerships, and a company that treats Australia as a market rather than an afterthought.

Is Claude safe enough for Australian businesses?

This is the question that comes up in almost every first conversation. Shortest possible answer: yes. Anthropic holds SOC 2 Type II, ISO 27001:2022 and ISO/IEC 42001:2023 (the new international standard specifically for AI management systems). Constitutional AI gives it a documented refusal policy. And on paid business plans, Anthropic doesn't train on your data by default.³

On the Privacy Act side: a data processing agreement is built into Anthropic's commercial terms, and if data residency matters for your APP 8 thinking, Claude is also available through AWS Bedrock in Sydney (ap-southeast-2). Same starting point Australian IT departments already accept for other cloud vendors. We've written a full breakdown in our guide to Claude and the Australian Privacy Act.

Which Claude models can you use in 2026?

Claude isn't one model. It's a family. And in 2026, the shelf is well stocked.

Claude Fable 5 and the Claude 5 family

Claude Fable 5 arrived on 9 June 2026 and is the most powerful model Anthropic has ever made generally available.⁶ It's one half of the Claude 5 family: Fable 5 and Mythos 5 share the same underlying model, but Fable is the public version with extra safety measures around dual-use capabilities, while Mythos goes only to approved organisations like cyber defenders and researchers.

Fable 5 is built for the long haul: tasks that run for hours, across millions of tokens, without losing the plot. We've written a full guide to Claude Fable 5 if you want the details.

Claude Opus 4.8

Opus is the powerful all-rounder and the most widely used heavyweight in production. It's built to hold long documents and whole codebases in its head at once (business plans come with a 200,000-token context window, roughly 150,000 words). Opus is what Mozilla used to review Firefox code (271 vulnerabilities found) and what NASA used to plan route options for Mars rovers.⁵

Claude Sonnet 5

Sonnet is the workhorse. Faster and cheaper than Opus but nearly as smart. For most Australian businesses, Sonnet is the obvious default. Not because it's the least. Because it's the most: good enough for 95 per cent of tasks without paying for the extreme.

Claude Haiku 4.5

Haiku is the fast, cheap one. Think of it as Claude when you need answers in milliseconds or need to chew through thousands of simpler tasks at once. Not for complex analysis. But perfect for chatbots, classification and real-time applications.

Claude Code and Claude Cowork

Beyond the models, Claude has become a product suite. Claude Code is an AI agent in the terminal that writes code, runs tests and fixes bugs. We've covered what Claude Code is and what it costs. Claude Cowork is the same idea with a graphical interface on Mac and Windows. Claude Design launched on 17 April 2026 and generates presentations, prototypes and marketing material from a text description.⁷ May 2026 added Claude for Small Business with ready-made automation templates and Claude for Legal built for law firms.

That's why we say Claude isn't a tool. It's a stack.

How does Claude AI work in practice?

An ordinary day at an Australian SME running Claude looks like this. The sales rep uploads an 80-page RFP and asks Claude to summarise the requirements. HR asks Claude to review an employment contract. The controller drops in a spreadsheet and asks for a variance analysis. The developer opens Claude Code in the terminal and lets it fix an old Python bug.

Everything runs through the same underlying models. But the interface adapts to the task.

Claude lives on the web (claude.ai), as a desktop app on Mac and Windows, as a mobile app, as a Microsoft 365 integration since March 2026, and as an API for developers building their own tools.⁷ And when you're ready to let it complete tasks rather than just answer questions, that's where Claude agents come in.

What does Claude AI cost?

Here's the price list, straight from Anthropic.⁸ Anthropic bills in US dollars; the AUD figures are approximate.

Plan	Price (USD)	Approx AUD	Who it's for
Free	$0	A$0	Test drive, limited usage
Pro	US$20/month (US$17 annual)	~A$30	One user, full toolkit
Max 5x	US$100/month	~A$150	Power users, 5x more usage
Max 20x	US$200/month	~A$300	Heavy daily use
Team Standard	US$20/seat/month annual (US$25 monthly)	~A$30	Teams from 2 seats
Team Premium	US$100/seat/month annual (US$125 monthly)	~A$150	Heavy usage, includes Claude Code
Enterprise	From US$20/seat + usage (self-serve), quote (sales-assisted)	From ~A$30/seat	50+ users or compliance needs

Those are Anthropic's prices. But there's one more option. With us at satori, managed Claude is A$850 setup plus A$15 per user per month on top of the licences (Standard A$42, Premium A$195). That includes one invoice in AUD, same-day support, a quarterly optimisation report and help with the rollout.

More detail in our Claude pricing guide for Australia, and in our breakdown of Claude Teams pricing for Australian businesses.

Is Claude better than ChatGPT?

Honestly? Not always. And that's not the wrong answer.

Here's the data. On SWE-Bench Pro, a hard software-engineering benchmark, Claude Fable 5 scores 80.3 per cent against GPT-5.5's 58.6.⁹ On long documents, Claude is the one built for the job: a 200,000-token context window on business plans and a model family trained to hold the thread across an entire contract stack. On image generation, Claude loses. It doesn't generate images. ChatGPT does.¹⁰

The honest conclusion is that for everyday work the two are close. The difference is what you do most of. Lots of text, code and analysis? Claude. Lots of images and voice? ChatGPT. Both? Then it's not strange at all. Plenty of big companies run both.

Read our Claude vs ChatGPT comparison for Australian businesses for the deeper analysis.

What does Claude AI mean for Australian businesses?

Three things.

Access to a frontier reasoning model with local backing. Anthropic's Sydney office opened in 2026, the company has signed a memorandum of understanding with the Australian government, and Australia is one of the heaviest per-capita users of Claude in the world.¹ This isn't a tool being trialled at the edge of the map. We've written more about Claude in Australia.

A realistic privacy story. A data processing agreement in the commercial terms, no training on business data by default, and a Sydney region option via AWS Bedrock. It doesn't solve everything automatically. But it's a much simpler starting position than most.

A platform to build the whole AI strategy around. That's why we at satori built our services on Claude. Not because it's the only choice. Because it's the choice most businesses land on, and specialising means we do it properly. Our guide to Claude for business walks through what that journey looks like.

The road from first chat to AI across the whole organisation usually runs through five steps. We call it the AI Ladder. Choose a platform. Explore deeply. Train on real work tasks. Structure your data. Automate. The training step is usually the one that decides whether the rest falls into place, and it's exactly what our Claude training covers.

How do you get started with Claude AI?

Three paths.

1. Sign up yourself. Go to claude.ai and create an account. Pro at US$20 a month. Or Team Standard from 2 seats. Works fine for one person or a small technical group.

2. Managed Claude. satori-claude is our managed service. We handle licences, support, training and quarterly reports. One invoice in AUD. A$850 setup plus A$15 per user per month on top of the licences. For businesses that want Claude without the procurement process.

3. A proper launch. satori-launch is our launch package at A$990 per user (from 8 users) that gets the whole team going at once. Webinar intro, a full-day workshop at your office and up to five skills rolled out across the organisation. And if you want a Claude consultant who walks with you the whole way (we're an AI consultant in Sydney, so yes, we do that), that's where the journey starts.

Ready for the next step? Get in touch and we'll take it from there.

Sources

Footnotes

Anthropic (2026). Anthropic Economic Index. https://www.anthropic.com/economic-index ↩ ↩²
Anthropic (2026). Anthropic News. https://www.anthropic.com/news ↩ ↩²
Claude by Anthropic (2026). Product Overview. https://claude.com/product/overview ↩ ↩²
Wikipedia (2026). Claude (language model). https://en.wikipedia.org/wiki/Claude_(language_model) ↩
Wikipedia (2026). Claude (language model). Major Partnerships and Investors. https://en.wikipedia.org/wiki/Claude_(language_model) ↩ ↩²
Anthropic (2026). Claude Fable 5 and Claude Mythos 5. https://www.anthropic.com/news/claude-fable-5-mythos-5 ↩
Anthropic Help Center (2026). Release notes. https://support.claude.com/en/articles/12138966-release-notes ↩ ↩²
Claude by Anthropic (2026). Plans & Pricing. https://claude.com/pricing ↩
Finout (2026). Claude Fable 5 and Mythos 5: Pricing, API Costs, and Benchmark Comparison vs Opus 4.8 and GPT-5.5. https://www.finout.io/blog/claude-fable-5-mythos-5-pricing-benchmarks ↩
Tech Insider (2026). ChatGPT vs Claude 2026: Full Comparison. https://tech-insider.org/claude-vs-chatgpt-2026/ ↩