AI Daily Briefing — 2026-06-08

OpenAI has unveiled a new security feature called Lockdown Mode, designed to offer enhanced protection against prompt injection attacks. These attacks occur when malicious instructions are hidden within webpages or other content sources, potentially tricking AI systems into exposing sensitive information. The feature aims to reduce the risk of data exfiltration for users handling confidential material.

Lockdown Mode works by restricting certain ChatGPT capabilities. It disables live web browsing, limiting access to cached content only, and blocks the retrieval and display of images from the web, though image generation remains available. Additionally, deep research and agent mode are turned off to minimize attack vectors. OpenAI acknowledges that even with this mode enabled, some vulnerabilities may persist, such as prompt injections in cached content or uploaded files.

The company emphasizes that Lockdown Mode is not for everyone. It targets individuals and organizations that manage sensitive data and require stricter safeguards against prompt injection risks. OpenAI is currently rolling out the feature to self-serve ChatGPT Business accounts and eligible personal accounts, marking a step forward in AI security.

The Trump administration is reportedly in discussions to acquire an equity stake in OpenAI, a move that could seed a proposed “Public Wealth Fund” aimed at distributing AI-driven profits directly to American citizens. President Donald Trump confirmed on Friday that he has spoken with AI companies about arrangements where “the American people can benefit from the success of AI,” though he did not name specific firms. CNBC later reported that the administration has been actively negotiating with OpenAI, with some of the equity potentially funding a fund the company recently proposed to share AI gains with the public.

According to Bloomberg, Trump told reporters aboard Air Force One that he has discussed “concepts where pieces could be given to the American public, where the American public essentially becomes a partner with the companies.” OpenAI CEO Sam Altman has reportedly floated the idea of government stakes in major AI firms since early 2025. The proposal aligns with Trump’s past intervention in the chip industry, where the government took a 10% stake in Intel last year.

The concept has also drawn support from progressive lawmakers. Senator Bernie Sanders this week proposed a one-time 50% tax on AI companies like OpenAI, Anthropic, and xAI, payable in stock, arguing it would “give the public a direct role in determining the future of this technology” and ensure AI wealth benefits everyone. However, former Trump AI and crypto czar David Sacks warned the plan could “accelerate the corporate-government fusion we’re already sliding toward,” while former Microsoft employee Dare Obasanjo suggested it lays “groundwork for a government bailout of OpenAI.”

OpenAI is preparing to launch a major update to ChatGPT in the coming weeks, transforming it into a so-called 'super app' that integrates coding tools and AI agents. The move, reported by the Financial Times, aims to help the company better compete with rival Anthropic, especially among business clients, while also pushing it closer to profitability ahead of a potential IPO.

By turning ChatGPT into a central gateway, OpenAI hopes to convert free users into paying customers for products like its Codex coding tool. Thibault Sottiaux, who leads OpenAI's core product and platform, described the vision as a personal agent capable of assisting users across all aspects of life, both personal and professional. One senior employee even declared, 'Chat is dead,' signaling a shift away from the traditional chatbot model.

This super app strategy is not new—reports of OpenAI's ambitions have circulated since last year. In March, The Wall Street Journal noted that the plan marks a major strategic pivot after the company launched several standalone products in 2025. OpenAI executives now say they are abandoning 'side quests' like the video generator Sora to focus on this unified platform.

Source: TechCrunch AI

Apple’s Worldwide Developers Conference (WWDC) 2026 kicks off Monday at 10 a.m. PT, with major updates expected for Siri and Apple Intelligence. The most anticipated change is a complete AI revamp of Siri, making it more conversational and capable of handling multi-step tasks across apps. According to reports, the new Siri will integrate Google’s Gemini technology and may debut as a standalone app to compete with chatbots like ChatGPT and Claude. Users could also gain the ability to set auto-delete timers for conversations.

Additional updates include an AI agent integration with the App Store, allowing users to delegate tasks like booking reservations or editing documents. The Camera app is expected to introduce a “Visual Intelligence” mode, replacing the previous Camera Control feature, and the Photos app will gain AI-powered tools such as intelligent scene recommendations and natural language editing. Image Playground will see upgrades for higher-quality image generation and a “describe a change” editing option.

The Wallet app may add a bill-splitting feature that scans receipts and generates payment requests, plus a “Create a Pass” option for digitizing physical tickets and cards. Across macOS, iPadOS, visionOS, watchOS, and tvOS, Apple plans to enhance AI-driven Siri experiences and introduce more stability updates. The event will be streamed live via Apple’s developer app, website, and YouTube channel.

President Donald Trump signed a new executive order on Tuesday focusing on artificial intelligence, marking a shift in the White House's approach to AI governance. The policy emphasizes innovation and security but has drawn criticism from both sides of the regulation debate. Key points include a voluntary review system where tech companies are asked to share frontier models with the government 30 days before release, no mandatory licensing requirements, and the creation of a dedicated AI cybersecurity clearinghouse to coordinate with the private sector. The order is a scaled-back version of a previous one Trump scrapped last month, which required 90 days' notice, but still represents a move toward stronger oversight.

Separately, defense-tech company Anduril has revealed new details about its augmented-reality headset for the military, developed in partnership with Meta. The device aims to allow soldiers to order drone strikes through eye-tracking and voice commands. Quay Barnett, a former Army Special Operations Command officer leading the project, envisions a cyborg-inspired system where drones and soldiers share information seamlessly and make decisions as one unit. The goal is to optimize the human as a weapons system, integrating vision and data for battlefield advantage.

These developments highlight the dual trajectory of AI: one path focused on civilian regulation and another on military application. The executive order seeks to balance innovation with security, while the smart glasses project pushes the boundaries of human-machine integration in warfare. Both efforts underscore the growing influence of AI in shaping policy and defense strategies.

This rewritten article is based on reporting from MIT Technology Review AI.