It is Monday, June 8, 2026, Apple held its annual developer conference, and we are not even going to bury the lede: Siri is now powered by Google. That is somehow only the second strangest thing that happened in AI today. Let us get into it.
Tim Cook’s last WWDC keynote just made Google the brain of Siri
Tim Cook took the stage at Apple Park this morning for his final WWDC keynote before handing the CEO role to hardware chief John Ternus on September 1. After two years of promised AI features that never shipped and a $250 million settlement over false advertising claims about Siri, Apple showed up today with a completely rebuilt assistant. The engine powering it belongs to Google.
Apple confirmed today that the new Siri runs on a custom 1.2-trillion-parameter Gemini model licensed from Google at approximately $1 billion per year. The rebuilt Siri arrives as a standalone app with a chatbot-style interface, a system-wide swipe gesture, Dynamic Island integration on iPhone 16 and newer, and personal context access across emails, photos, messages, calendar, and files. On-screen awareness and multi-step cross-app task execution were demonstrated on stage — precisely the features Apple promised at WWDC 2024 and never delivered.
The detail that most coverage is underplaying: iOS 27, iPadOS 27, and macOS 27 will let users pick their preferred AI model for Apple Intelligence. Gemini is the default. Claude and ChatGPT are selectable alternatives, each with a distinct voice so you know which one answered. This is the most significant AI platform policy Apple has made in a decade, and it quietly turns the iPhone into a multi-provider AI marketplace with two billion active devices as the distribution channel.
Apple’s privacy argument: the heavy Gemini processing runs on Apple’s own Private Cloud Compute servers, not Google’s infrastructure. Whether regulators and security researchers will accept that distinction is a question that will be answered in the months ahead.
Read the live coverage: Tom’s Guide WWDC 2026 live blog
Read the Siri and Gemini breakdown: TechTimes on the WWDC keynote | Fast Company on what it means for Apple’s strategy
Donald Trump and Bernie Sanders are agreeing on AI policy, and Sam Altman flew to Capitol Hill to encourage both of them
This is real. Donald Trump said Friday that the US government may take direct equity stakes in AI companies like OpenAI, Anthropic, and xAI so that “the American people essentially become a partner.” He described it as “a beautiful thing.” Bernie Sanders, meanwhile, published a New York Times op-ed calling for the American AI Sovereign Wealth Fund Act — a one-time 50% tax on equity from the major AI labs with board representation for Washington. The two men are not saying the same thing, but they are saying it in the same direction.
In between those two statements, Sam Altman flew to Washington and met privately with Bernie Sanders for nearly an hour at the senator’s request — after Sanders announced the ownership plan. According to people with knowledge of the conversation, Altman told Sanders he also wants the public to hold equity in AI companies, just not at 50%. He has been pitching some version of this idea to Trump since early 2025.
The underlying pressure driving all three men toward the same position: tech layoffs in 2026 have already surpassed 142,000. Software developer employment for workers aged 22 to 25 has fallen roughly 20% from its 2024 peak. OpenAI’s own internal documents project a $14 billion loss in 2026 alone — and it is spending that money building something that is visibly displacing workers. The politics of that math are not sustainable, and apparently everyone in the room can see it.
Read the coverage: Fortune on Trump’s equity stake comments | Washington Times on the Altman-Sanders meeting | Fortune on Sanders’ original wealth fund proposal
The Pentagon fired its AI provider for refusing to help build autonomous weapons, and the replacement auditions are not going well
The story that keeps getting stranger: the US Defense Department is still running replacement auditions for Anthropic’s Claude, which was designated a “supply-chain risk” by Defense Secretary Pete Hegseth on February 27 after Anthropic refused to remove two specific safety guardrails — no mass surveillance on Americans, and no fully autonomous weapons systems that target people without human input. Claude had been the Pentagon’s primary AI provider, deeply embedded in the Maven Smart System used for classified operations.
Twenty-five military “power users” spread across five global commands are now evaluating models from OpenAI, Google, and xAI’s Grok against identical prompts. The evaluation is running on a platform separate from Maven specifically because replacing the deeply integrated Claude deployment is described by contractors as requiring 12 to 18 months and significant productivity losses. Anthropic is challenging the supply-chain-risk designation in court, arguing it could cost the company billions. Pentagon talks with Anthropic remain formally suspended because of that lawsuit.
The situation has produced a genuinely strange dynamic: multiple government departments are intentionally delaying the transition, hoping for a negotiated resolution before the six-month deadline. Pentagon power users have told officials that Claude outperforms the alternatives being tested. The military fired its AI for being too principled and is now discovering, three months later, that the replacements are not as good.
Read the latest: TechTimes: The Pentagon is racing to replace Claude for being “too safe for war” | Bloomberg’s original reporting on the replacement tests | CryptoBriefing on the full timeline
The SpaceX roadshow is live and it is unlike any IPO ever attempted
SpaceX kicked off its investor roadshow last Thursday, June 4, targeting a $75 billion raise at a $1.75 trillion valuation — which would make it the largest initial public offering in history by a significant margin. Pricing is June 11. Trading begins June 12 on Nasdaq under the ticker SPCX.
The unusual feature is who gets to participate. SpaceX reserved 30% of the offering for retail investors — three times the standard proportion for a company of this size — accessible through Fidelity with a minimum account balance of just $2,000. Elon Musk will not sell a single share. The dual-class structure gives him full voting control after listing. The company disclosed 2025 revenue of $18.7 billion, driven primarily by Starlink’s satellite broadband business, which now serves more than nine million subscribers.
The AI angle: the SpaceX offering is the first of three massive AI-adjacent IPOs expected in the second half of 2026. Anthropic has confidentially filed its S-1. OpenAI is expected to follow. All three are competing for the same institutional capital pool, which means the price SpaceX sets this week will function as the benchmark for both Anthropic and OpenAI when their turns come. If SPCX prices strong, Anthropic and OpenAI benefit. If it disappoints, both companies face a more skeptical market.
Read the coverage: Full roadshow timeline and retail access details | Polymarket odds on the IPO completing in June
Also worth reading today
- Claude’s chatbot market share grew 306% in a single quarter, from 203 million to 824 million web visits between January and April 2026. ChatGPT still leads at 54.7% of worldwide web visits but is down from 76.5% in early 2025. Gemini is second at 27.4%. (Build Fast With AI June 8 roundup)
- Anthropic launched a formal $100 million enterprise partner program this week, called the Claude Partner Hub, structured to reward firms that implement Claude at scale. Tier standings are reviewed twice yearly, and firms can only move down at an annual December 31 review. (Build Fast With AI June 8 roundup)
- The EU AI Act general-purpose AI obligations deadline is now 55 days out. Any company with a frontier model deployed in Europe needs to have governance documentation, transparency reports, and systemic-risk assessments ready by early August. The clock is running. (Build Fast With AI June 8 roundup)
That is your AI Monday. Siri works for Google now, the far left and the far right want to nationalize your chatbot, and the US military is stuck in a breakup it cannot afford to finish. See you tomorrow.