Anthropic’s Public Benefit Fund: An AI UBI Plan? Plus OpenAI Aurora and OpenCode

Key takeaways

• Anthropic is committing an initial 5% of its shares to a Public Benefit Fund, with up to 2% more for each 50% valuation rise, subject to board approval
• The video says the wider pledge comes to roughly 20% of shares; the UBI link is speculation, not an announced plan
• OpenAI image models called Aurora Red, Green and Blue Alpha were reportedly seen on Arena, with mixed early results
• Partners reportedly have a second Fable 5.5 checkpoint, and a Claude personal finance feature may be on the way
• OpenCode reported a breach affecting about 1.4% of console users who had external providers connected
• Smaller items: Gemini Omni 1.5 with 15-second video, Odyssey 3 world model, a Claude Code config mod and MiniMax M3.1 open weights
• For a small business, the useful actions are about account security, careful use of finance integrations and calm planning for automation

Every weekend brings a pile of AI headlines, and most of them do not change how you run your business. This one has a couple that might. Anthropic has put a large slice of its company into a fund for the public, and a popular developer tool has had a serious security incident.

This article sums up the WorldofAI roundup in plain English, explains what each story means for a small business, and separates official announcements from leaks and speculation.

What is Anthropic’s Public Benefit Fund?

It is an organisation Anthropic has set up to support the public as AI changes the economy, funded with company shares. Anthropic describes it on its own news page, so this part is official.

Using a reported valuation of $960 billion, the video puts the first 5% at around $48 billion of equity. That is the presenter’s sum on a reported figure, not a cash amount in a bank. The presenter also calls it partly a marketing move, which is a fair point to keep in mind.

Is it really a UBI plan?

Not as far as anyone has said. The UBI idea comes from one phrase in the announcement: direct distribution to specific populations. That could mean cash payments, help for people whose jobs are automated, or something else entirely.

Our take: the more useful signal is that a leading AI company is planning openly for job disruption. You do not need to guess at AGI to act on that. A sensible, calm plan for your own team is worth more than any headline, and our guide How to Plan for AI Automation in a Small Team Without Panic walks through one.

What is new with Fable 5.5 and Claude?

Partners reportedly have a second, stronger Fable 5.5 checkpoint, and Claude may gain a personal finance feature. Both are reports, not launches.

• Fable 5.5: the video shows a 3D Superman animation built in Blender with Python code, credited to the new checkpoint. We covered the earlier leaks in Claude Fable 5.5 Leaks: Is It Really “Baby AGI”? Plus GLM 5.5 and GPT-6 Astra Pro.
• Claude personal finance: internal findings reportedly show a Plaid-style bank connection, spending analysis and personal dashboards. It was apparently codenamed “Claude Money” before being renamed.

A bank connection is the kind of feature where you should slow down. Before you link any AI tool to your accounts, run through our bank account safety checklist.

What are OpenAI’s Aurora models?

They appear to be three image generation models being tested anonymously on Arena: Aurora Red Alpha, Green Alpha and Blue Alpha. OpenAI has not announced them.

Early results are mixed. Some examples look decent, while others reportedly did badly in Arena comparisons. The presenter also shares a wider frustration that OpenAI has been leaning on usage resets and small updates since its GPT-6 launch. As part of its 28-day shipping pledge, day 13 was a usage reset and day 14 added the ability to edit earlier messages in the desktop app.

What happened at OpenCode?

OpenCode, an AI coding tool, reported a security breach linked to a large-scale model distillation attack. According to the company, about 1.4% of console users were affected, specifically those who had external AI providers connected. Affected users have been notified.

• Attackers reportedly targeted OpenCode’s systems for nearly 48 hours
• The AI models the team used to investigate started refusing to help partway through, possibly because of their safety rules
• Access to OpenAI’s Daybreak Blue cyber-defence setup reportedly helped
• OpenAI says it worked with OpenCode to restore access and refund credits used during the attack, and suggests teams request cyber access before an incident, not during one

What else was in the roundup?

The video also mentions DeepSeek V4.1 Flash 1017, which we tested in DeepSeek V4.1 Flash 1017 Tested: How a Cheap Open Model Stacks Up. If you use Claude Code heavily, our post on cutting token waste pairs well with the config mod idea.

What is confirmed and what is not?

For a method to check this kind of news yourself, see How to Verify AI Model Leak Rumours Before You Switch Tools.

What should a small business do with this news?

1. Rotate API keys you have stored in AI tools, and set spending caps with each provider.
2. Hold off on linking bank accounts to AI assistants until you have checked permissions and data handling.
3. Write a simple automation plan for your team, so changes happen on your terms.
4. Ignore model rumours until a launch is official and you can test it on your own tasks.

Frequently asked questions

The bottom line

Anthropic’s Public Benefit Fund is real, large and worth watching, but the UBI angle is speculation for now. The more practical lessons from this roundup are closer to home: keep your API keys tidy, be careful before giving AI access to your money, and plan for automation calmly rather than waiting for the next headline. Our two practical guides are a good place to start.

Related posts

• Claude Fable 5.5 Leaks: Is It Really “Baby AGI”? Plus GLM 5.5 and GPT-6 Astra Pro
• How to Verify AI Model Leak Rumours Before You Switch Tools
• Self-Replicating AI Agent Swarms: What They Mean for Your Website

Garry Knight

I'm Garry Knight, the person behind Prodify Digital. I write about email list building, email marketing, SEO, AI search and the tools that connect them. My aim is to make online marketing easier to understand, so creators and small business owners can make informed decisions about building an audience and keeping people engaged. Here you'll find straightforward guides and product reviews that explain what something does, where it fits and which limitations matter. The focus is on clear explanations and useful next steps—not hype, shortcuts or promises of easy earnings.

See Full Bio