VeryCodedly This Week: Anthropic Pulls Fable 5, SpaceX Makes History, and Countries Team Up for Tech

Another week, another rundown. Here's this week's.

SpaceX Makes History

The company went public on Thursday, with stock opening at $150, above the $135 IPO price, and closing at a 19% gain. After raising $75 billion, it became the largest IPO in history, closing at more than double the last record holder, with a $2.1 trillion market cap. But…

Only investors who got shares at $135 actually made that 19% gain. If you bought at the opening price of $150, you're up only 7.3%. If you bought just three minutes later at 11:49 AM, you're up 1%. Most investors who bought later in the day actually lost money because the stock peaked at $176.52 early and then dropped. Anyway, the benchmark has been set. Anthropic, you're up next.

Anthropic Pulls Its Best AI

Well, that didn't last long.

After staying up for less than a week, Anthropic cut off access to Fable 5 and Mythos 5. Worldwide.

Why: US government orders. The government gave them verbal evidence of a jailbreak method that could bypass the models' safety safeguards, basically calling it a national security risk. Anthropic reviewed it and says the capability shown is "widely available from other models (including OpenAI's GPT-5.5)". The company's statement is worth reading, they're clearly annoyed.

Their words: "If this standard was applied across the industry, we believe it would essentially halt all new model deployments".

India's Tech Show-and-Tell

India brought 120 deep-tech startups to France this week, to talk about AI, quantum computing, semiconductors, space tech, biotech, plus lots more, and over 500 international investors showed up to watch and take notes. The country already provides the manpower for tech hubs around the world, so building the tech too is a no brainer. Now they want the world to know.

India is also GitHub's fastest-growing developer hub, with 3.5 million developers joining from India in 2023 alone. The country now has over 17 million developers on GitHub. Go India!

UK-Japan $24B Tech Deal

UK and Japan just locked in $24 billion for tech and infrastructure. AI, chips, quantum, semiconductors. Major corporations like Hitachi and Rolls-Royce are also investing, and the money also includes billions for UK offshore wind projects.

AI really is creating all kinds of partnerships.

German Court Makes Google Liable for AI Answers

Google's AI Overviews falsely claimed that two Munich-based publishers were involved in scams, subscription traps, and "dubious business practices". The AI hallucinated information and made up connections that weren't in any of the linked sources. The publishers were not satisfied with Google's response to their cease-and-desist letter so they sued.

The verdict from the Regional Court of Munich was a temporary injunction barring Google from repeating those false claims and 80% coverage of the legal costs. This is a preliminary injunction from a regional court, not a final judgment, so Google can appeal. This could happen to any AI.

AWS Requires Data Sharing for Anthropic Models

Before AWS's AI platform (Amazon Bedrock) added Fable 5 and Mythos 5, prompts and completions stayed inside AWS's boundary, so model providers (like Anthropic) never saw the data. Now the only way to use Fable 5 or Mythos 5 on Bedrock is to opt into "provider_data_share" mode, which means your data leaves AWS's security boundary and is shared with Anthropic, retained for 30 days, and subject to human review.

By flipping this switch, Anthropic becomes a "sub-processor" of your data, which is both a technical and legal change. Companies using AWS Bedrock may need to revise privacy policies, and re-evaluate whether they can legally use these new models at all.

That's all for Things That Happened this week. More stuff will happen, so see you next week.