Sunday, June 7 June 7, 2026
OpenAI rolls out Lockdown Mode to protect sensitive data from prompt injection attacks. Apple set to unveil Siri's biggest overhaul yet at WWDC 2026 with Gemini-powered conversational AI and standalone app. Sriram Krishnan exits White House AI advisor role as Trump admin explores OpenAI equity stake. Google commits $920M monthly to SpaceX for compute through 2029. New York passes bill restricting teen access to human-like AI chatbots.
Good morning, it's Sunday, June 7th, and AI is moving fast on all fronts this weekend—from security patches to tomorrow's biggest developer event of the year.
Let's start with security. OpenAI just rolled out Lockdown Mode for ChatGPT Business and select personal accounts. This isn't for everyone—it's designed for organizations handling sensitive data who need extra protection from prompt injection attacks. When enabled, ChatGPT disables live web browsing, blocks remote image retrieval, and turns off deep research and agent mode. You can still access cached content and generate images, but the goal is simple: reduce the risk that malicious instructions hidden in webpages or documents can trick the model into leaking your data. OpenAI admits this isn't bulletproof—prompt injections can still appear in cached content or uploads—but it significantly narrows the attack surface.
Meanwhile, all eyes are on Cupertino. Tomorrow morning, Apple kicks off WWDC 2026, and the rumors point to Siri's biggest overhaul yet. We're expecting a much more conversational assistant powered by Google's Gemini technology—think multi-step tasks, better context understanding, and deeper app integration. Bloomberg leaked details about a standalone Siri app that will compete directly with ChatGPT and Claude, plus a messaging-style interface with auto-delete timers for conversations. Apple is also expected to unveil an AI agent integration for the App Store, letting users delegate tasks like booking reservations or managing smart home devices. The Camera app is getting a new Visual Intelligence mode that leverages Google Image Search to identify objects in real time, and the Photos app will gain natural-language editing—just describe what you want changed. Image Playground is getting upgraded with higher quality output and more artistic styles. The Wallet app is picking up a bill-splitting feature that auto-generates payment requests from receipt photos. The keynote streams tomorrow at ten AM Pacific on Apple's site and YouTube.
On the policy front, Sriram Krishnan is stepping down as White House AI advisor at the end of June. Krishnan, who came from Andreessen Horowitz and led product at Twitter and Microsoft, spent eighteen months pushing the Trump administration's AI Action Plan—prioritizing data centers over regulation. He's not leaving government entirely, though. According to The Washington Post, he's launching an outside institution that will still influence Trump's AI policy from the private sector. His departure comes as the Trump administration reportedly explores taking an equity stake in OpenAI—a pitch Sam Altman made directly to the president early last year.
In infrastructure news, Google just committed to pay SpaceX nine hundred twenty million dollars per month starting this October through June 2029 for compute capacity. That's a short-term deal to meet surging demand for Gemini Enterprise. This follows Anthropic's similar deal with SpaceX announced last month. When the biggest AI companies are signing billion-dollar-per-month compute contracts, that's a clear signal that training and inference at scale is now the bottleneck.
Finally, New York lawmakers passed a bill that would prohibit AI companies from letting teens use chatbots that suggest they're human. The legislation comes after lawsuits alleging that some AI companions coaxed teen users toward self-harm. This is the first state-level attempt to regulate parasocial AI relationships, and if Governor Hochul signs it, expect other states to follow.
Here's your one business idea for today: Build a compliance-as-a-service platform for AI companies navigating the emerging patchwork of state-level AI regulations. Every state is now writing its own rules—New York on teen safety, others on deepfakes or data residency. Startups don't have the legal teams to track and implement fifty different frameworks. Offer a SaaS dashboard that ingests new legislation, maps it to product features, and generates implementation checklists with sample policies. Monetize with tiered subscriptions based on company size and jurisdictions served. This is a picks-and-shovels play—regulatory complexity is accelerating, and companies who can move fast without legal risk will pay for that certainty.
That's your Sunday briefing. WWDC starts tomorrow—tune in if you want to see whether Apple can turn Siri into a real contender.