Woman Joins Lawsuit Against Musk's AI

Welcome to OHMO AI. A woman has joined a lawsuit against Elon Musk's xAI, alleging Grok was used to turn a photo of her at 11 into at least 7,000 explicit images. The claim is about the guardrails that were supposed to stop exactly that.

The rest of today runs on numbers. Anthropic put a quarter on the table that dwarfs the same quarter last year. Alibaba's open models cleared at least 3 billion downloads. And Anthropic says future Claude models will write with a watermark inside the text. The money and the damage are arriving at once.

In today's newsletter:

OHMO AI: A woman identified in the case as Jane Doe 4 has joined a lawsuit that three Tennessee teenagers filed against Elon Musk's xAI over the part Grok allegedly played in creating child sexual abuse material. She alleges her stepfather fed Grok a photo taken when she was 11 and used it to make more than 7,000 explicit images of her. Her words for it: "It is taking everyday life and turning it into child sexual abuse."

The details:

Why it matters: Image filters are built to catch the obvious prompt from a stranger. This claim isn't about a stranger, it's about a family photo and someone who already had it, which is the scenario nobody designs the safety rails around. If a court decides the company that ships the tool carries a duty of care for what comes out of it, every consumer AI product gets rebuilt around that answer.

Global hiring is changing fast. From AI's impact on HR to evolving compliance requirements and international expansion strategies, the rules are constantly shifting.

Oyster's events bring together HR leaders, founders, operators, and global employment experts to discuss what's working now—and what's coming next.

Whether you're hiring internationally today or planning for tomorrow, you'll walk away with practical insights you can actually use.

Explore Hiring Events

OHMO AI: Anthropic put preliminary revenue for its latest completed quarter at more than $11.5 billion, against $787 million in the same quarter of 2025. The company also reported positive adjusted operating income for its second quarter. That's the Claude maker turning into a different size of business inside a year.

The details:

Why it matters: Income changes who a company answers to. Burning investor money buys the freedom to be picky about what you sell and to whom, and paying customers at this scale come with expectations attached. Every careful decision Anthropic makes from here has a price tag someone can see.

OHMO AI: Alibaba's open-weight AI models have been pulled down more than 3 billion times over the past six months, on figures from the open-source AI hub Hugging Face. That puts it ahead of Google's 418 million in 2026 and Meta's 227 million. Open-weight means anyone can take the model onto their own hardware and run it there instead of calling a company's servers.

The details:

Why it matters: Downloads aren't revenue, but they're where habits set. Developers build on whatever's already on their machine, and the model that gets built on quietly becomes the engine inside apps nobody thinks of as Chinese. The centre of gravity in open AI moved, and no government decided it.

Dictate acceptance criteria and reproductions inside Cursor or Warp. Wispr Flow auto-tags file names, preserves syntax, and gives you paste-ready text in seconds. 4x faster than typing.

Try Wispr Flow free

OHMO AI: Anthropic has set out how watermarking will work in what Claude writes. Future models will produce text that carries a watermark, a signal built to show how likely it is that Claude had a hand in a piece of writing rather than to settle it outright. The company published the mechanics itself.

The details:

Why it matters: Schools, editors and employers have been arguing about AI written text with nothing but instinct and bad detector tools. A mark the model carries itself changes the quality of that argument, and it changes who gets believed. The risk is that a probability gets treated as proof by people who need a clean answer, and a student or a writer pays for the difference.

See three budget futures at once instead of one forecast

A single forecast gives a decision-maker one number to react to. Three scenarios side by side let them see where the pressure actually lands and ask better questions on the spot. Ten minutes, no spreadsheet required.

Copy this, swap in the bracketed parts, and paste it in.

You don't need exact figures to start, rough percentages by category are enough for Claude to build comparable visuals across futures. Naming a real variable, like a grant that might disappear, anchors the scenarios in something concrete instead of abstract percentages. Asking for the toggle between dollars and percentages matters because a proportional shift and an absolute cut tell different stories, and a decision-maker needs both. Once you see the shape of it, iterate with constraints like "hold rent fixed and show me where the pressure actually lands" to test the assumptions that matter to your situation. This is built for speed and clarity in a room, not for final numbers.

Every newsletter, we feature how a reader is using AI to work smarter, save time, or make life easier.

Today's reader is Marta K. from Gdansk, Poland:

“I run a bakery here that supplies about twenty cafes. The baking was never the hard part. The hard part is that every cafe orders by text message at whatever hour suits them, and then half of them change it at 6am. I used to sit with a notebook at four in the morning adding it all up. Now I paste the whole mess of messages in and I get one list by product and one delivery run per driver, and it flags anyone who ordered less than they normally do so I can ring them. Two weeks ago it caught that a cafe had gone quiet for eleven days. They'd switched supplier and nobody told me. I got them back. My accountant asked if I'd hired someone.”

Want to be in the next newsletter? Tell us how you use AI here.

AI Tools: A curated list of the AI tools worth knowing. 

Instagram: The AI page they pray you’ll never find.

Facebook: For the ones who never left Facebook.