AI News — June 07, 2026

Security

OpenAI Unveils Lockdown Mode to Protect Sensitive Data from Prompt Injection Attacks
OpenAI has introduced a new 'Lockdown Mode' designed to safeguard sensitive user data from prompt injection attacks. While the feature reduces the likelihood of sensitive data being exposed, ChatGPT may still carry some vulnerability to such attacks even with the mode enabled.

Products

OpenAI Still Developing a 'Super App' as Senior Employee Declares 'Chat is Dead'
OpenAI is continuing work on an ambitious 'super app' that goes beyond conversational chat interfaces. A senior OpenAI employee signaled a major shift in product philosophy, suggesting the company envisions a more expansive, integrated application experience.

Models

What to Expect from WWDC 2026: Siri Revamp and Apple Intelligence Updates
Apple's Worldwide Developers Conference 2026 is approaching, with a highly anticipated overhaul of Siri and significant updates to Apple Intelligence expected to be front and center. The event is seen as a critical moment for Apple to reassert its AI ambitions amid growing competition.

Society

AI Content Creators Are Becoming Increasingly Difficult to Spot
AI-generated influencers and content creators have grown significantly more convincing, making it harder for audiences to distinguish them from real humans. The trend raises growing concerns about authenticity, transparency, and the potential for misinformation in the creator economy.

Tooling

Notion Restores Access to Anthropic After Service Disruption
Notion experienced a service disruption affecting its integration with Anthropic's AI, which has since been resolved. The incident drew notable attention online, with Notion's head of product expressing surprise at the level of social media engagement around the outage.