What Happened In The Room This Month. The June 2026 Fishing Report.

We Built a Korean-Language Awareness Video, a Planting Tool, New Websites, and Solved a Financial Puzzle That Had Been Sitting for Months. All Live. All in the Room.

June was one of the most varied months we have had in a long time. Six very different groups. Six very different industries. And in every single room, the same thing happened. The moment AI stopped being theoretical, everything changed.

At a workshop for the Pacific Northwest's largest beverage manufacturer, a team member uploaded two large financial datasets that had been causing headaches for months. Four minutes later, AI solved it. Someone in the back quietly said: 'Oh shit, it actually solved it.' That is the moment we live for. Not because the AI worked. Because a room full of leaders suddenly saw what was possible.

For Alzheimer's Orange County, we built a fully localized, Korean-language awareness video, created live in the room. Not a rough draft. A finished, usable piece of content their team could deploy immediately. Not a writing assistant. A production partner.

Across two sessions with CEO peer groups in Vancouver and Edmonton, we built new websites from scratch, created financial dashboards from raw data, and worked through real business challenges brought into the room by participants. Every output was functional. Every output was built live, with the audience watching every step.

For an Entrepreneurs' Organization chapter, we built a customer-facing planting tool for a seed company, live on stage. For infrastructure CFOs, we focused on financial modeling and strategic decision support. Different tools. Same result. Leaders left with something they could use the same day.

None of this was prepared in advance. No staged demos. Everything built from scratch, in the room, with the audience watching. The feedback collected through SpeakFeed continues to tell the same story: 100 percent found the session valuable, 100 percent would attend again, and 97 percent or more rated sessions as engaging, relevant, and interactive. One attendee wrote: 'He made this new thing a lot less scary. I changed from AI is too scary to even try, to Wait, I wonder what this could do for me.'

Anthropic Launched Its Most Powerful Model Ever. Two Days Later, the US Government Shut It Down.

There has been no bigger story in AI this month than what happened with Anthropic's Fable 5. It launched on June 10, 2026, as the most capable model Anthropic had ever released. Two days later, on June 12, the US government ordered it offline.

Fable 5 was not a routine model upgrade. Anthropic described it as a significant leap in capability, particularly for complex, multi-step reasoning and professional work. It launched alongside a companion model called Mythos 5. Both were immediately available to Claude subscribers.

On June 12, the US government issued an export control directive ordering Anthropic to suspend all access to Fable 5 and Mythos 5 for any foreign national, whether inside or outside the United States. The practical effect was that Anthropic had to disable both models for all customers globally to ensure compliance. The government's stated concern was a potential jailbreak. Anthropic reviewed the technique and found the vulnerabilities were minor, already known, and replicable using other publicly available models. In their public statement, Anthropic was direct: they disagree with the decision, believe it sets a dangerous precedent, and are working to restore access.

Access to all other Anthropic models, including Claude Sonnet, Haiku, and Opus, was not affected. If your team uses Claude for day-to-day work, nothing changed for you. The bigger implication is about the direction of AI regulation. This is the first time the US government has used export control authority to pull a commercial AI model from the market. It will not be the last. Anthropic stated: 'If this standard was applied across the industry, we believe it would essentially halt all new model deployments for all frontier model providers.'

ChatGPT's Image Generation Is Now Genuinely Stunning. And Claude Just Launched a Design Tool That Changes How You Build Things.

Two tools caught our attention this month. Both are worth understanding. Both are immediately useful.

ChatGPT Images 2.0, built natively into GPT-4o, is a completely rebuilt image generation system. For years, AI image generation had a persistent weakness: text inside images was unreliable. That has changed. Text now renders cleanly and accurately. The system supports multilingual text in images, advanced visual reasoning, and a new thinking mode for complex creative requests. What makes this practically useful is the editing workflow. You can describe changes in plain language, select a specific part of an image to modify, or adjust the overall style, all within the same conversation. No plugins. No separate tools. For teams that create marketing materials, presentations, or social content, this is worth exploring today.

Anthropic launched Claude Design as part of its Anthropic Labs research preview, available now to Pro, Max, Team, and Enterprise subscribers. The concept is straightforward. You describe what you want in a chat interface on the left. Claude builds it on a live canvas on the right. You iterate through conversation and inline comments until it is right. Then you export or share it.

What Claude Design can produce is not a rough sketch. It generates working designs, interactive prototypes, presentations, and one-pagers. If you upload your existing codebase or design files, it automatically extracts your brand's colors, typography, and component patterns and applies them to everything it builds. That is the feature that separates it from other tools. It does not just generate something that looks good. It generates something that looks like yours. That is the difference between a demo and a tool you can actually use.