OpenAI and Anthropic this week: GPT-Red, Fable 5 plan changes, and free Claude for teachers (Week…

2026-07-19 · Source: AI on Medium · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation, Corporate Strategy & Leadership · Depth: Intermediate, medium

Summary

OpenAI and Anthropic released numerous updates in Week 29, 2026. OpenAI introduced GPT-Red, an automated red teamer for prompt injection vulnerabilities, and is reportedly developing a screenless smart speaker. ChatGPT returned to WhatsApp in the EEA, gained unified search, and increased custom instructions to 5,000 characters. OpenAI also published GPT-Live usage limits and a guide for managing AI investments. Anthropic made Claude Fable 5 standard in Max and Team Premium plans at 50% of limits, offering a one-time \$100 credit. They launched free premium Claude access for verified US K-12 educators, committed \$10 million CAD to Canadian AI research, and enhanced Claude Code artifacts with public sharing and multiplayer editing. Anthropic also made HIPAA configuration self-serve for organizations and published research on Claude's values and agentic misalignment.

Key takeaway

For AI Product Managers evaluating new tools or strategizing AI adoption, you should assess the implications of these rapid updates. Consider integrating GPT-Red for enhanced security testing and leverage Claude Fable 5's expanded access for advanced applications. Evaluate your AI investments using OpenAI's "Useful Intelligence per Dollar" metric to ensure cost-effective and impactful deployments.

Key insights

Leading AI labs are rapidly advancing model capabilities, safety, and accessibility across diverse applications.

Principles

• Automated red teaming enhances AI security at scale.
• AI value alignment can be quantified across four axes.
• "Useful Intelligence per Dollar" guides AI investment.

Method

GPT-Red uses adversarial self-play to find prompt injection vulnerabilities. Claude's values were compressed from 300K+ conversations into four axes. Claude Code's /code-review uses subagents for deep analysis.

In practice

• Utilize ChatGPT's 5,000-character custom instructions.
• US K-12 educators can access free premium Claude.
• Evaluate AI investments using "Useful Intelligence per Dollar."

Topics

• OpenAI
• Anthropic
• AI Safety
• Large Language Models
• Prompt Engineering
• AI Investment
• Educational AI

Best for: CTO, VP of Engineering/Data, Director of AI/ML, AI Product Manager, Tech Journalist

Related on AIssential

• 😸 Anthropic and OpenAI Both Dropped Their Best AI Models On The Same Day — Leading AI developers are advancing models towards autonomous job execution, emphasizing either broad utility or deep specialization.
• Anthropic has caught up to OpenAI in image understanding — Claude Fable 5 and GPT-5.5 demonstrate significant progress in image understanding, yet general intelligence and complex geometric reasoning remain elusive.
• Anthropic Claude Fable 5 — Anthropic's new Claude Fable 5 offers superior agentic capabilities but introduces controversial, silent performance limitations for specific research tasks.
• Anthropic Steals OpenAI's Crown — The AI landscape is rapidly diversifying with advanced models, specialized tools, and integrated agent platforms.
• The AI Industry as You Know It Died Today — The US government's intervention in frontier AI model releases signals a shift towards centralized, restricted access, mirroring a "Manhattan Project for AI" approach.
• Bernie Sanders endorses Cori Bush’s comeback bid — Anthropic released Fable 5, a powerful AI model with robust guardrails, balancing advanced capabilities with safety for public use.

See Counsel's argued verdicts on the open AI decisions leaders are weighing →

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AI on Medium.