What's New in AI: Apple's Gemini Deal — 15 June 2026

The biggest names in AI spent this week pairing off and cashing in. Apple handed the brains of Siri to Google's Gemini, ChatGPT quietly passed a billion monthly users, and OpenAI joined Anthropic in the queue for a blockbuster public listing — all while Washington and Western Australia set fresh rules and budgets for the technology. Here is what mattered, and why it should be on your radar.

Apple bets Siri's future on Google's Gemini

At its WWDC 2026 keynote on 8 June, Apple finally unveiled Siri AI, a long-delayed overhaul of its assistant — and confirmed it runs on Google's Gemini models rather than anything built in-house. Reports put the arrangement at roughly US$1 billion a year for a custom, 1.2-trillion-parameter version of Gemini. The new Siri can hold a back-and-forth conversation, act on what is on your screen, summarise documents, and pull tasks out of a photo or schedule. It rolls out across iPhone, iPad, Mac, Apple Watch and Vision Pro this spring — but not evenly. Apple is blaming the European Union's Digital Markets Act for indefinitely delaying Siri AI on iPhones and iPads in Europe, where it will initially appear only on Mac and Vision Pro.

Why it matters: Two of the fiercest rivals in technology are now quietly dependent on one another, a telling sign of just how hard frontier models are to build from scratch. For everyone else, the deal cements conversational AI as a default phone feature rather than a novelty. And the European carve-out is a preview of a more fragmented world, where the same product ships with different capabilities depending on local law. You can read the full keynote rundown at Engadget.

ChatGPT crosses a billion users — but the pack is closing

New Sensor Tower data released on 12 June shows ChatGPT reached a billion monthly users in roughly three and a half years, the fastest any consumer app has hit that mark. The catch is that rivals are growing faster from smaller bases: Anthropic's Claude jumped about 640% year on year and Meta AI surged 973%, against ChatGPT's still-healthy 62%. Adoption is broadening at work, too — Boston Consulting Group found that 74% of frontline workers now use AI regularly, with more than 40% of frequent users saying they save the equivalent of a full workday each week.

Why it matters: AI assistants have crossed from early-adopter curiosity to everyday infrastructure, even as public scepticism rises in parallel. For business owners, the takeaway is less about which chatbot "wins" and more that AI fluency is fast becoming a baseline expectation among staff and customers alike. If you are still sizing up where these tools fit, our running library of AI tools and guides is a practical place to start. The milestone was first reported by CNBC.

OpenAI joins Anthropic in the IPO queue

On 8–9 June, OpenAI confirmed it had confidentially filed a draft S-1 with the US Securities and Exchange Commission — the first formal step toward an initial public offering that could value the company near US$1 trillion, with Goldman Sachs and Morgan Stanley managing the deal and a listing possible later in 2026. It lands just days after Anthropic's own filing, which means two of the leading AI labs are now in the pipeline at once. The mood music came from SpaceX, whose record-breaking Nasdaq debut on 12 June closed up about 25% on day one and set an early benchmark for the mega-listings to come.

Why it matters: Going public forces these labs to open their books, justify enormous spending, and prove the business — not just the technology — is durable. For customers, more financial transparency is welcome; it makes it easier to judge which providers will still be standing in five years. It is a notable escalation from Anthropic's filing, which we unpacked a fortnight ago. The OpenAI move was confirmed by Fortune.

Washington writes new rules for frontier AI

A White House executive order, Promoting Advanced Artificial Intelligence Innovation and Security, directs federal agencies to stand up a voluntary framework for "covered frontier models" — including a process for developers to give the government early access to powerful new models for up to 30 days before wider release — and to prioritise AI in cybersecurity. Agencies are working to tight 30- and 60-day timelines, with key deliverables due 2 July and 1 August. At the state level, Colorado's AI Act — the first comprehensive state AI law in the US — takes effect on 30 June, imposing impact assessments and anti-discrimination duties on "consequential" automated decisions in areas such as hiring, lending and housing.

Why it matters: Regulation is shifting from drafting to enforcement, and it increasingly lands on the businesses using AI, not just the labs building it. If you deploy AI in hiring, credit or customer decisions, governance and documentation are quickly becoming part of the product rather than an afterthought. The order is published on the White House site, with a useful state-law tracker from Cooley.

Western Australia puts money behind public-sector AI

On 10 June, WA Innovation Minister Stephen Dawson announced a A$10 million AI Investment Fund and a new Public Sector AI Centre of Excellence, alongside the state's Digital Strategy 2026–2030. The fund is designed to attract industry and university partners to pilot and scale AI projects with measurable productivity or service-delivery benefits, paired with a procurement mechanism meant to give agencies easier, less-duplicated access to AI suppliers.

Why it matters: It is a concrete signal that Australian governments are moving from AI talk to AI budgets — and the centralised, outcomes-first framing mirrors what already works in the private sector. For local businesses, public-sector demand opens real opportunities, but it rewards those who can show results rather than slideware. Turning a promising model into something dependable is exactly the gap we help organisations close when we build and deploy custom AI systems. The announcement was covered by Startup News.

Quick hits

• Google's agentic "auto browse" arrives in Chrome on Android in late June, letting Gemini complete multi-step chores such as booking appointments and finding in-stock items — while asking for confirmation before purchases or social posts. Google
• Anthropic launched a US$100 million Claude Partner Network, adding a tiered Services Track and a Partner Hub for finding implementation specialists; it says more than 40,000 companies have already applied. Anthropic
• Observability firm Coralogix raised a US$200 million Series F at a roughly US$1.6 billion valuation, a reminder that infrastructure-layer AI is still attracting big cheques. Tech Startups
• Beyond Siri, Apple Intelligence is spreading across default apps — organising Safari tabs by topic, fixing weak passwords in one tap, and adding smart replies in Messages and Mail. Engadget

Looking ahead

Watch for more frontier-model news as the US executive order's first deadlines arrive in early July, and for Colorado's AI Act to become a live compliance test on 30 June. With OpenAI and Anthropic both circling public markets, the next few weeks could reset how the whole industry talks about money — and how much it is willing to reveal.

That is the wrap for this week. The pace is not letting up, so check back next week for the next instalment of What's New in AI — and let us know if there is a story you would like us to dig into.