1,200 per cent. That was the median year-on-year rise in AI sessions across 115 Australian businesses between September 2024 and September 2025. OpenAI rolled out ChatGPT-5 free to all users while Google’s Gemini showed broad data collection in app-store disclosures, Anthropic filed to register in Australia, and xAI’s backers kept insisting Grok leads on raw capability. The real fight is now about reach, privacy posture and managed deployment, and Australian integrators are already packaging Claude for security-conscious customers. Anthropic says it will set out a local strategy in the first half of 2026, a milestone that could reshape commercial uptake here.

Volume and commerce are pulling in different directions: ChatGPT dominates referrals, but other platforms are converting users into purchases at higher rates.

The claim that matters for Australia isn't only who builds the smartest model, but who pairs capability with governance and local delivery. That's the central thesis of this moment. OpenAI’s free, global rollout of ChatGPT-5 gives it unmatched reach, but Anthropic’s corporate moves and Google's data posture have shifted procurement calculus for regulated organisations and enterprise buyers.

Capability, and the case for GPT-5

OpenAI announced ChatGPT-5 as available to all users and its chief executive, Sam Altman, described the model as "clearly a model that's generally intelligent," while also cautioning it doesn't continuously learn in deployment. OpenAI told journalists the model delivers large capability gains across coding, writing and healthcare workflows and promoted "vibe-coding" as a defining use case going forward.

Public materials and early adopter commentary characterised GPT-5 as more competent and consistently accurate than prior releases. One independent early-access commentator described it as feeling like "a PhD-level expert" on many topics. Those capability claims are the core of OpenAI’s leadership argument: broad capability, free access and massive user numbers as the strategic wedge.

Countering that claim, xAI’s backers including Elon Musk publicly disputed the assertion that GPT-5 is the absolute leader. Musk wrote on X that his Grok 4 Heavy model "was smarter" than ChatGPT-5. XAI’s positioning continues to emphasise capability where its backers see advantage in model design and engineering choices.

Mastheads

At the same time, the market metrics paint a split story. An analysis of 115 Australian businesses from September 2024 to September 2025 found AI sessions rose a median 1,200% year-on-year, yet AI-referred purchase rates sat at 2.9% versus 6.0% for organic traffic. Within that dataset, Perplexity and Google’s Gemini converted three to six times better than ChatGPT on e-commerce outcomes, even as ChatGPT remained the dominant referrer in volume. Those results suggest reach doesn't automatically equal commercial return.

Procurement decisions in Australia are also shifting on data posture. Google’s app-store disclosures, summarised in a consumer privacy review, list broad categories of data Gemini may access including browsing history, contacts, emails, photos and precise location. That openness can be valuable for some consumer scenarios but creates a privacy trade-off for organisations and privacy-sensitive customers.

Anthropic has taken a different route. The company filed to register Anthropic Australia in December, listing California-based legal and governance officers on its filings, and has signalled plans to outline its local strategy in the first half of 2026. Local channel activity has already started: Think Technology Australia launched a Claude-focused, security-led managed service in mid-April aimed at small and mid-market customers. Think Technology’s operations director, Mungo Bright, said Anthropic’s partner programme and training were key to packaging the product for clients. Bright told customers Claude excels at tasks such as financial-data analysis and that managed services must combine platform expertise with ongoing governance and iteration.

That combination of partner enablement, local operation and governance is precisely the differentiator enterprise buyers name when they talk about uptake. Anthropic and its channel partners argue the fight for enterprise adoption will be decided by ongoing governance and integration rather than raw model benchmark scores.

Those three fault lines, capability, privacy posture and managed deployment, are why industry voices disagree about who's "leading." OpenAI points to GPT-5 as a general-purpose capability leader and stresses free, global availability and user numbers. XAI’s backers push capability claims for Grok. Anthropic and its partners focus on safety, partner enablement and managed, local deployments. Google’s privacy disclosures place Gemini at the other end of the data-collection spectrum, which matters for procurement choices in regulated sectors.

Historical context matters. OpenAI has expanded commercial and infrastructure commitments globally over the past two years and secured large-scale data-centre deals that underpin its distribution. Anthropic, founded in 2021 by former OpenAI employees, has scaled fundraising and product development quickly and now positions Claude as a principal competitor. In Australia both companies have made concrete moves to establish local presence while integrators adapt platforms to regulated enterprise needs.

Availability and pricing remain a practical concern. OpenAI marketed ChatGPT-5 as a free rollout at announcement.

Anthropic’s Australian subsidiary registration is public and local Claude-focused managed services went live in mid-April through partners, though the company had not publicly named a country head or fully specified commercial terms in its filings. Gemini’s privacy disclosures are visible in app-store listings. The commerce study covers activity through September 2025 and predates some recent product updates, so measured outcomes may shift as models and distribution features change.

Related Articles

Anthropic plans to outline its Australian strategy in the first half of 2026. That milestone will show whether managed services, partner training and local governance can turn Claude’s enterprise pitch into measurable market share here.

This article was created with AI assistance.