Forbes Newsletters

Meta’s AI app is filled with people’s private conversations with AI.

Forbes

Presented by

Welcome back to The Prompt. 

In every major technological revolution, people lose their jobs. AI will be no exception, Anthropic cofounder and CEO Dario Amodei said recently. He predicts that AI could result in mass elimination of entry-level jobs across fields like tech, law, finance and consulting and increase the unemployment rate to 10 to 20 percent in the coming years. Other leaders have opined on AI killing jobs too. Geoffrey Hinton, touted as the Godfather of AI, said paralegals or customer service reps are on the chopping block. LinkedIn cofounder Reid Hoffman posted on X: “Some AI industry leaders are predicting white-collar bloodbaths,” suggesting that fresh graduates “AI optimize” themselves by understanding and adopting AI in creative ways.  

Let’s get into the headlines.

Rashi Shrivastava Reporter

Follow me on Forbes.com

BIG PLAYS
OpenAI’s bromance with Microsoft appears to be fracturing. The latest symptom: OpenAI executives have discussed accusing Microsoft of violating antitrust law and engaging in anticompetitive behaviour during the partnership, according to the Wall Street Journal. The two companies are split on the terms of OpenAI’s $3 billion acquisition of AI coding startup Windsurf as well as how much stake Microsoft would have if OpenAI converts into a for-profit public benefit corporation, per WSJ. Microsoft was pivotal during OpenAI’s early days, but their relationship has frayed as OpenAI has rapidly grown. 
ETHICS + LAW
The National Association for the Advancement of Colored People has filed an intent to sue Elon Musk’s AI company, xAI for contributing to air pollution in predominantly Black communities in Memphis through its supercomputer, Forbes reported. Last year, Musk built a massive “gigafactory of compute” in Memphis in record speed to power his AI company. In the process, he sidelined council members from the project, made officials sign NDAs and installed “quick and dirty” natural gas turbines to get the data center up and running as quickly as possible. 
Presented by ServiceNow
Bad CRM was then. This is ServiceNow.
On the ServiceNow AI Platform, AI is built in—not bolted on. Supercharged with AI agents, it brings your organization together. Because CRM for the AI era isn’t about managing customers—it's about seamlessly connecting the entire business to delight them.

Put AI To Work
DATA DILEMMAS
Hollywood powerhouses Disney and Universal are suing popular AI image generator Midjourney for copyright infringement, alleging that the company is a “copyright free-rider and a bottomless pit of plagiarism,” and that it has made copies of iconic characters like Frozen’s Elsa and the Minions from Despicable Me without their permission by feeding copyrighted works into its models. The lawsuit is the latest in a litany of litigation filed by creatives against a host of AI startups for training their AI models on intellectual property without consent or compensation. 
Scale AI CEO Alexandr Wang  Drew Angerer/Getty Images
DEEP DIVE
OpenAI Has Been Winding Down Its Work With Scale
Read Article
Scale AI’s shock $15 billion deal to sell a 49% stake to Meta, with its CEO Alexandr Wang leaving the startup he founded to lead a new AI lab at the tech giant, has sparked fears that the startup’s $14 billion-valued data labeling business could lose clients to competitors.

At play is the concern that Scale, which has been the dominant player in labeling data to help major tech companies and AI startups train their models, could share details about the types of data that leading AI players have used to build their most cutting edge tech with Meta.

As one former Scale employee told Forbes: “They all want to cut Scale off now. Scale as a business, once it becomes part of Meta, entirely collapses.”

OpenAI, one of its most prominent customers, has already been winding down its work with Scale, according to four sources with knowledge of its business; two sources said this has been happening for months and OpenAI has been vetting potential new partners. Scale AI initially declined to comment, but after publication, spokesperson Joe Osborne denied that OpenAI has pulled back its spending with the company.

The surprise deal, which would value the company at $28 billion, left people inside Scale scrambling and confused, one former employee said. Some staff are concerned over what visibility Meta might have into past projects, even though most contracts with the company stipulate data be deleted after projects are completed, the person said.

Forbes has learned that Meta intends to pay Scale at least $450 million a year for five years for its AI products, or more than half of its annual AI spend, whichever is less, according to a person who worked on the deal. That commitment from Meta would make up a major part of Scale’s revenue. Last year, Scale generated $870 million in revenue, largely driven by demand for its data labelling services. Scale declined to comment on the figures and Meta did not respond to a request for comment. 

Scale’s smaller rivals are already jostling for position, and are welcoming any clients worried about conflicts of interest when it comes to data and privacy. “We're already seeing a huge influx of demand from customers that are phasing out of Scale AI,” said Brendan Foody, CEO of $2 billion-valued Mercor. Invisible Technologies cofounder Francis Pedraza told Forbes his company is committed to staying independent. Startup Turing, which already supplies data for model training to OpenAI, Anthropic and Google, sees the deal as an opportunity to be “Switzerland” and become an impartial distributor of data to the frontier AI labs. CEO Jonathan Sidharth told Forbes that “clients want to work with someone neutral who could support all the labs equally.” One investor who has backed a Scale competitor said the deal would create new avenues for other firms to “capture the open space left by Scale AI.”

Read the full story on Forbes. 

YOUR WEEKLY DEMO
What if AI sat on both sides of the (virtual) bargaining table? A recent study tested two AI agents — systems that can autonomously carry out specific tasks— to see how well they could negotiate with one another. The study found that stronger agents, with access to more powerful AI models, were able to exploit weaker ones to get a better deal. As consumers use AI to make purchasing decisions and businesses use AI to communicate with customers, AI-mediated deal making might not be far from reality — but those with better models would have the upper hand. 
MODEL BEHAVIOR
Meta’s standalone AI app is teeming with people’s personal conversations with AI, Business Insider reported. From medical questions about vitamin supplements to personal legal advice about getting fired, the app’s public feed is filled with intimate interactions that people seem to have accidentally posted for the wider public to see. One such post included a person’s email address and phone number. In response to the story, Meta added a pop-up warning to remind people that their posts are public. But the social media giant’s AI problems extend far beyond exposing people’s private chats: its models have struggled to keep up with rivals.
MORE FROM FORBES