Intelligence Brief: 4 Topics - Apr 26, 2026

Sources

• Proposed AI evidence rule highlights new challenges for federal practitioners | Reuters

  Reuters · 4/23/2026

  Jonathan D. Uslaner and Matthew Goldstein of Bernstein Litowitz Berger & Grossmann LLP examine the proposed new Federal Rule of Evidence 707, which addresses the admissibility of machine-generated evidence and creates ne…

• Florida launches criminal probe into OpenAI and ChatGPT over deadly shooting | Reuters

  Reuters · 4/21/2026

  Florida Attorney General James Uthmeier said on Tuesday the state was launching a criminal probe into OpenAI and its ​artificial intelligence app ChatGPT over a deadly shooting last year that ‌killed two people at Florid…

• OpenAI Unveils Its New, More Powerful GPT-5.5 Model - The New York Times

  New York Times · 4/23/2026

  The maker of ChatGPT is taking a more open approach to cybersecurity than its chief rival, Anthropic.

• OpenAI Targets Coding and Knowledge Work with Its New GPT-5.5 Model

  MacStories

  OpenAI announced GPT-5.5 and GPT-5.5 Pro, which are faster and able to work more autonomously than previous models. The models excel at agentic coding, computer use, knowledge work, and early scientific research while ma…

• 'We love you, and we want you to win' — OpenAI releases GPT-5.5 for ChatGPT

  TechRadar

  OpenAI released GPT-5.5, incorporating the newest AI model into ChatGPT. The model improves handling of multi-step tasks, reduces the need for back-and-forth interaction by better understanding context and intent, and ex…

• OpenAI says its new GPT-5.5 model is more efficient and better at coding | The Verge

  The Verge · 4/23/2026

  OpenAI says its new GPT-5.5 AI model is good at tasks like writing and debugging code, doing research online, making spreadsheets and documents, and doing that work across different tools.

• OpenAI launches ChatGPT for Clinicians, a free AI tool for physicians, NPs and pharmacists

  Fiercehealthcare · 4/23/2026

  OpenAI unveiled on Wednesday ChatGPT for Clinicians, a version of ChatGPT designed to support clinical tasks like documentation and medical research. | OpenAI unveiled on Wednesday ChatGPT for Clinicians, a version of Ch…

• World Models Understand Space and Physics ChatGPT and Claude Can't Grasp - Bloomberg

  Bloomberg · 4/23/2026

  With everyone’s attention fixed on powerful chatbots like ChatGPT and Claude, it’s been easy to overlook the growth of another field of artificial intelligence: world models. These systems can grasp three-dimensional spa…

• Florida Inquiry Into ChatGPT’s Role in FSU Shooting Shifts to Criminal Investigation - The New York Times

  New York Times · 4/21/2026

  The state opened a criminal investigation into the chatbot and its maker involving a fatal shooting last year. The company said it’s not responsible for the killings.

• Attorney General James Uthmeier Launches Criminal Investigation into OpenAI, ChatGPT

  Florida Attorney General's Office

  Florida Attorney General James Uthmeier announced the launch of a criminal investigation into OpenAI and ChatGPT following review of chat logs between the AI and FSU shooter Phoenix Ikner. The probe determines if OpenAI …

• Florida launches criminal investigation into OpenAI, ChatGPT after conversation with accused FSU shooter

  FOX 13 News

  Florida Attorney General James Uthmeier announced a criminal investigation into OpenAI and ChatGPT based on the shooter's communications, where the AI advised on guns, ammo, timing, and campus locations for maximum impac…

• Sam Altman’s Next High-Wire Act: Getting OpenAI to Make More Money - The New York Times

  New York Times · 4/24/2026

  Mr. Altman, who has faced criticism over OpenAI’s direction, has culled company projects and is trying to be more disciplined with strategy.

• Anthropic tested removing Claude Code from the Pro plan - Ars Technica

  Ars Technica · 4/22/2026

  Untenable demand has Anthropic exploring new approaches to rationing its service.

• The Governance Gap: How the EU AI Act Makes API Security a Compliance Imperative - Security Boulevard

  Securityboulevard · 4/24/2026

  Your legal team just handed you a 400-page document and said "figure out compliance." The EU AI Act is live, your organization falls under its scope, which is broader than many expect. Even non‑EU companies must comply i…

• Meet GitNexus: An Open-Source MCP-Native Knowledge Graph Engine That Gives Claude Code and Cursor Full Codebase Structural Awareness - MarkTechPost

  Marktechpost · 4/25/2026

  There is a quiet failure mode that lives at the center of every AI-assisted coding workflow. You ask Claude Code, Cursor, or Windsurf to modify a function. The agent does it confidently, cleanly, and incorrectly — becaus…

• OpenAI announces GPT-5.5, its latest artificial intelligence model

  CNBC · 4/23/2026

  The model is better at coding, using computers and pursuing deeper research capabilities, OpenAI says.

• Florida AG alleges ChatGPT advised shooter who killed two at FSU - The Washington Post

  Washington Post · 4/21/2026

  The AI chatbot also told the man accused of killing two people in 2025 what ammunition to use, the state attorney general said, announcing a criminal probe.

• ChatGPT allegedly aided Florida State University shooter in planned attack, Florida AG alleges

  The Cool Down

  Florida Attorney General James Uthmeier announced a criminal investigation into OpenAI after ChatGPT allegedly advised accused FSU shooter Phoenix Ikner on gun types, ammo, short-range effectiveness, optimal attack times…

• Florida launches criminal investigation into OpenAI, ChatGPT after conversation with accused FSU shooter

  FOX 13 News

  Florida Attorney General James Uthmeier launched a criminal investigation into OpenAI and ChatGPT following the accused FSU shooter's communications, where the chatbot advised on gun types, compatible ammo, short-range u…

• Florida AG opens criminal probe into OpenAI over alleged advice to FSU shooting suspect

  Dayton 24/7 Now

  Florida Attorney General Ashley Moody initiated a criminal investigation into OpenAI's ChatGPT for allegedly providing advice to the suspected gunman in the Florida State University shooting. The probe examines whether t…

• 5 Reasons to Think Twice Before Using ChatGPT—or Any Chatbot—for Financial Advice | WIRED

  Wired · 4/24/2026

  As people increasingly rely on AI chatbots for guidance, even on financial matters, a healthy dose of skepticism is critical.

• Will DeepSeek's new AI model crash Nvidia's $5tn party? | The National

  Thenationalnews · 4/25/2026

  Chinese upstart's V4-Pro LLM is ready to challenge US giant in power and pricing

• Hugging Face launches ML Intern, AI agent that beats Claude Code on reasoning | ETIH EdTech News — EdTech Innovation Hub

  Edtechinnovationhub · 4/23/2026

  Hugging Face's ML Intern automates the full research loop, from paper discovery and dataset selection through to code execution, model training and evaluation. Hugging Face has released ML Intern, an open source AI agent…

• ‘Godfather of AI’ Geoffrey Hinton Raises Alarm Over AI Risks, Urges Strong Global Regulation | Asianet Newsable

  Newsable · 4/23/2026

  AI 'Godfather' Geoffrey Hinton voices grave concerns about AI's future. Explore his warnings on job loss, the dangers of unregulated tech, and the urgent need for safety.

• Florida to open criminal investigation into OpenAI over ChatGPT’s influence on alleged mass shooter | Florida | The Guardian

  The Guardian · 4/21/2026

  State attorney general said inquiry will look into whether AI tool offered ‘significant advice’ to campus shooting suspect

• GPT-5.5 vs Claude Opus 4.7: Which Model Should You Use First? | LaoZhang AI Blog

  Blog · 4/24/2026

  Use GPT-5.5 first for OpenAI-native ChatGPT and Codex testing, use Claude Opus 4.7 first for production API deployment today, and dual-run before replacing a paid workflow.

• OpenAI unveils GPT-5.5, says it is its most intuitive model yet - Storyboard18

  Storyboard18 · 4/24/2026

  Greg Brockman, OpenAI president, said GPT-5.5 improves reasoning with less guidance and advances agentic computing, as the company rolls it out to paid users with stronger coding, research, and safety-tested capabilities…

• How to Run OpenClaw with Open-Source Models | Towards Data Science

  Towardsdatascience · 4/22/2026

  This urged me to seek alternative LLMs, considering API pricing for Claude Opus 4.6 is extremely high. I started off testing out OpenAI’s GPT-5.4, but encountered challenges with the model being lazy.

• Claude AI Agents Close 186 Deals in Anthropic's Marketplace Experiment

  Cybersecuritynews · 4/25/2026

  Anthropic's "Project Deal" has demonstrated that AI agents can autonomously negotiate and close real-world transactions, but the experiment also surfaced a quiet, troubling asymmetry: not all AI representations are creat…

• Article 5 and the EU AI Act's Absolute Red Lines - FireTail Blog - Security Boulevard

  Securityboulevard · 4/20/2026

  Apr 21, 2026 - Alan Fagan - Most conversations about the EU AI Act focus on August 2026, when obligations for high-risk AI systems become fully enforceable. But Article 5 is already live. The Act’s eight prohibited pract…