🤖 OpenAI releases GPT-5.2

Good Morning! OpenAI just went into full code-red mode and dropped GPT-5.2, a model that wasn’t expected this soon and is already outperforming expectations across coding, math, reasoning, and long-context intelligence. Plus, today I’ll show you how to access and run pre-built n8n automations instantly.

In today’s AI newsletter:

• OpenAI Launches GPT-5.2

• Google Unveils Gemini Deep Research Agent

• Trump Signs Executive Order Targeting State AI Laws

• Access 7000+ Pre-Built n8n Automations in Minutes

• 4 new AI tools worth trying

AI MODELS

🤖 OpenAI Launches GPT-5.2

OpenAI rolled out GPT-5.2, its newest frontier model designed for professional work, long-running agents, and high-stakes reasoning. According to OpenAI, the model brings major upgrades over GPT-5.1 in intelligence, reliability, and long-context accuracy, and was pushed out early after Google’s Gemini 3 showed strong benchmark gains.

• Sets new SOTA on SWE-Bench Pro (55.6%) and SWE-Bench Verified (80.0%), with major gains in multi-step agentic coding

• Near-perfect long-context accuracy on MRCRv2 “needle-in-haystack” tasks up

  to 256k tokens

• New highs on ARC-AGI-1/2, FrontierMath, and GPQA Diamond

• Error rates halved on chart reasoning and GUI tasks (CharXiv, ScreenSpot-Pro)

• Tool calling hits 98.7% accuracy on Tau2-bench Telecom — huge for autonomous agent workflows

GPT-5.2 shows that OpenAI is still very much in the race, and may even be ahead on several high-value benchmarks. With major boosts in coding, reasoning, and long-context reliability, this model could set the standard for next-gen agents, enterprise workflows, and complex multi-step tasks.

AI NEWS

🧠 Google Unveils Gemini Deep Research Agent

Google launched a reimagined version of Gemini Deep Research, powered by its flagship Gemini 3 Pro model. What used to be a tool for generating research reports is now a full-blown agentic research system that developers can embed directly into their apps.

• Built on Gemini 3 Pro, Google’s “most factual” model designed to reduce hallucinations

• Developers can use the new Interactions API to embed deep research capabilities

• Handles massive context windows and complex, multi-step reasoning tasks

• Already used for due diligence, drug toxicity analysis, and large-scale information synthesis

To prove its improvements, Google released a new benchmark called DeepSearchQA, plus results on existing tests like Humanity’s Last Exam and BrowserComp, where its agent performed strongly across the board.

Google is preparing for a future where you don’t “Google search” anything, your AI agent does. And the timing of this announcement wasn’t subtle: OpenAI dropped GPT-5.2 the same day, claiming superior performance across key benchmarks. The AI rivalry is escalating fast, and agents are the new battleground.

AI REGULATIONS

⚖️ Trump Signs Executive Order Targeting State AI Laws

President Trump has signed a sweeping executive order pushing to limit or overwrite state-level AI regulations, aiming to centralize AI policy under the federal government. The order can’t directly erase state laws, but it empowers federal agencies to challenge them, pressure states financially, and discourage future AI restrictions.

• Creates a new AI Litigation Task Force to sue states with “burdensome” AI laws

• Directs the FTC to outline when state rules that “alter truthful AI outputs” are federally preempted

• Commerce Department must identify states whose laws may cost them federal broadband funding

• FCC tasked with creating nationwide AI disclosure standards that supersede conflicting state rules

This EO tries to shift AI regulation power from states to Washington, bypassing Congress after multiple failed attempts to pass a national moratorium on state AI laws. Critics say it hands enormous influence to AI czar David Sacks and pushes the federal government into murky legal waters. Supporters argue it prevents a messy “50 states, 50 rulebooks” scenario that could slow down US AI innovation.

HOW TO AI

🗂️ Access 7000+ Pre-Built n8n Workflows

In this tutorial, you’ll learn how to browse, download, and import fully built n8n workflows from the public library with over 7000+ workflows.

🧰 Who is This For

• Creators automating social media and content workflows

• Engineers and data teams working with APIs and databases

• Solo founders optimizing daily operations

• Anyone who wants to install production-ready n8n workflows in minutes

STEP 1: Access the Tool

Open your browser and go to n8nworkflow. On the homepage, you’ll see a large search bar at the center. Type in the workflow you are looking for, such as Instagram, Resume Analysis, or PostgreSQL.

If you prefer browsing, use the category tabs at the top of the screen. Categories include Marketing, Engineering, AI, Productivity, and more. Each category opens a gallery of curated workflows you can explore.

STEP 2: Open a Workflow

Click on any workflow card that appears in the results. This brings you to its detail page.
Here you’ll find a section explaining how the workflow works and a full preview of the node diagram.

Scroll through the diagram to confirm the workflow includes the automations you need, such as API calls, webhooks, AI steps, or database actions.

STEP 3: Download the Workfow

On the right side of the workflow detail page, you’ll see a panel with a yellow button labeled Download Workflow.

Click this button to download a .json file. This file contains the exact node configuration and logic required for n8n to recreate the automation on your system.

STEP 4: Import the Workflow into Your n8n Instance

Open your n8n dashboard, whether it’s the desktop app, cloud version, or a self-hosted instance.

Create a new workflow by clicking Add Workflow or open a blank canvas.
In the top-right menu, click the three-dot icon. Choose Import from File and select the .json file you downloaded.

n8n will instantly load the entire automation onto your canvas with all nodes and connections intact.

Disney and OpenAI sign a three-year deal to bring 200+ Disney characters to Sora; Disney will make a $1B investment in OpenAI and become a major OpenAI customer

OpenAI's Fidji Simo says ChatGPT's “adult mode” is expected to launch in Q1 2026, says OpenAI wants to get better at age prediction before launching the feature.

xAI partners with El Salvador to deploy Grok across 5,000+ public schools, reaching 1M+ students in an “AI-powered education program” over the next two years.

Time names “the Architects of AI” as its 2025 Person of the Year and takes a look at AI's effect on the economy, geopolitics, and how we interact with the world

🧠 GPT-5.2: OpenAI’s faster, smarter model with improved reasoning

🧮 Nomos 1: Nous Research’s advanced AI for math reasoning

🌱 Seedream 4.5: ByteDance’s improved AI for creating and editing images

🧠 Purpose: An AI mentor that gives personalized guidance anytime

THAT’S IT FOR TODAY

Thanks for making it to the end! I put my heart into every email I send, I hope you are enjoying it. Let me know your thoughts so I can make the next one even better!

See you tomorrow :)

- Dr. Alvaro Cintas