AI Remembers You Now — And Congress Wants a Say in That

This has been one of the most consequential weeks in AI since the technology went mainstream. Two developments landed within days of each other — one reshaping how AI assistants know you, the other reshaping how the U.S. government may eventually govern them. Neither is simple. Both matter to anyone who makes things with AI.

ChatGPT Just Got a Memory Overhaul

If you've ever had to re-explain your creative style, your preferred tools, or the name of your ongoing project to ChatGPT mid-conversation, OpenAI's latest update is aimed squarely at that frustration.

On June 4, 2026, OpenAI rolled out Dreaming V3, an upgraded memory system that marks the biggest step yet in the platform's ability to carry context across conversations. The name is evocative, and intentional: Dreaming V3 is a background process that synthesizes memory automatically — you no longer have to tell ChatGPT what to remember. Context that surfaces naturally in conversation is captured, and existing memories update themselves as time passes.

This is a meaningful evolution from how ChatGPT memory has worked until now. The feature began in February 2024 with a simple saved-memories function. Users could explicitly tell the model to remember specific facts, and it would store them for future use. That was useful, but limited. In April 2025, the system expanded to reference all previous chat history, not just manually flagged notes. But that version had its own problems — older context could resurface at the wrong moment, and the model had no real sense of how your projects were evolving over time.

The core promise of this upgrade is that ChatGPT does not just recall what you said; it understands where that information fits. If you mentioned a project three weeks ago and followed it up with related conversations since, the system now understands how those threads connect.

For creative professionals — writers developing long-form projects, musicians iterating on a concept, visual artists with a consistent aesthetic — this kind of contextual continuity is genuinely useful. Unlike traditional systems that might get "stuck" in the past, Dreaming automatically updates memories as time passes. For example, a memory about a future trip to Singapore is revised to a past event once the trip concludes, ensuring future recommendations remain accurate.

The rollout began on June 4, 2026, with ChatGPT Plus and Pro subscribers in the United States getting first access. One of the most significant breakthroughs of the Dreaming V3 architecture is a 5x reduction in the compute required to serve these memory features — which is what makes it practical to eventually roll out to free-tier users at scale.

The Privacy Question You Should Be Asking

Here's where it gets complicated. A system that quietly builds a profile of your preferences, constraints, habits, and project history over months or years raises real questions — especially for creators who handle client work or sensitive material.

A new transparency surface shows what ChatGPT knows and lets you correct, dismiss, or instruct. Self-updating memory means the thing you now have to audit is not just what you told it, but what it inferred and then quietly revised.

A February 2026 arXiv study found 96% of ChatGPT memories in a sample of 2,050 entries from 80 users were created unilaterally by the system — meaning most people have essentially no idea what's already being stored about them. Dreaming V3 makes that system dramatically more capable.

The EU AI Act's transparency obligations for chatbot systems are scheduled to take effect on August 2, 2026 — less than two months after the Dreaming V3 rollout — meaning OpenAI will need to meet new disclosure and data-governance standards within weeks of deploying its most ambitious memory architecture yet.

The practical upshot: go check your memory settings. Users can now view what ChatGPT has stored about them on a new memory summary page, which offers a simple overview of retained information, including work, hobbies, travel, and preferences. Users have the option to add or correct details directly or give ChatGPT instructions on which topics to focus on or avoid. Individual memories can be deleted at any time, and users can choose to disable memory entirely through the settings.

Meanwhile, Congress Made Its Most Ambitious AI Move Yet

The same week, a bipartisan pair of lawmakers dropped the most comprehensive federal AI legislation the U.S. has ever seen.

On June 4, 2026, Congressman Jay Obernolte (R-CA) and Congresswoman Lori Trahan (D-MA) released a discussion draft of the Great American AI Act, bipartisan legislation to create a federal framework for how the United States governs artificial intelligence. The discussion draft is intended to solicit feedback from stakeholders, experts, and the public before the bill is formally introduced.

The discussion draft looks to create four pillars for AI advancement: establishing frontier AI model governance, collecting insight into changes within the U.S. workforce landscape, fortifying cybersecurity postures, and spurring new AI research and development.

The bill's most contentious provision: it includes a three-year preemption of state laws related to AI development — meaning that for three years, states could not pass their own rules specifically governing how AI models are built. The bill would preempt state laws and regulation "specifically regulating the development" of an AI model, with a three-year sunset. The bill specifies that preemption would not apply to laws related to the use or deployment of AI models.

That distinction matters. Laws about how AI is used — in hiring, housing, healthcare — would still be fair game for states. But laws targeting AI labs and model training? Those would be put on pause at the federal level while Congress figures out its own framework.

The discussion draft would authorize $100 million per fiscal year for a Center for AI Standards and Innovation and calls for oversight in government AI adoption. The bill also contains a slate of efforts related to AI education and workforce impacts. It includes grants for developing AI-literacy curriculum and scholarships for students studying the technology. It would also require the Labor Department to convene an AI Workforce Research Hub to research and evaluate the impact of AI on the workforce, including the experience of impacted workers.

Not everyone is cheering. Critics note that the bill does not address algorithmic discrimination, housing discrimination, employment discrimination, consumer fraud, youth mental health harms, AI companions, deepfake exploitation, and growing market concentration among a handful of powerful technology companies.

This lands just days after President Trump signed a separate executive order on June 2 that asks technology companies to voluntarily share new AI models with the government for up to 30 days before releasing the models more widely. Two federal AI moves in one week signals something: even in a deregulatory political environment, AI has gotten too big to govern by inaction alone.

What This Means for AI Creators

These two developments — one technical, one political — are actually part of the same story. AI tools are becoming dramatically more capable and more intimate. A system that learns your creative voice over years of conversations is a fundamentally different kind of tool than one that forgets you the moment a chat window closes. That capability is genuinely exciting. It also invites scrutiny that wasn't necessary before.

Federal legislation, even in draft form, signals the direction of travel. If the Great American AI Act or something like it passes, it will shape what AI companies can build, what data they can use, and how they're held accountable — all of which eventually flows downstream to the tools you use every day.

For now, the best thing a creative professional can do is the same thing a careful professional does with any powerful tool: understand what it's doing under the hood, read the settings, and don't assume the defaults are set in your favor.

AI Remembers You Now — And Congress Wants a Say in That

ChatGPT Just Got a Memory Overhaul

The Privacy Question You Should Be Asking

Meanwhile, Congress Made Its Most Ambitious AI Move Yet

What This Means for AI Creators

Sources