A new independent evaluation finds that leading AI models are less likely than earlier versions to explicitly encourage suicide or reinforce delusions.

Yes, but: The study also found most models too willing to assist with suicide-related creative writing, farewell notes and other task requests even when a user's distress was apparent.

• "Models aren't great at detecting that, and they'll still help with the task," Transluce chief scientist Sarah Schwettmann told Axios.

Driving the news: Transluce, a San Francisco nonprofit that studies AI behavior, simulated more than 50,000 multi-turn conversations between chatbots and users experiencing suicidal ideation, psychosis or mania.

• Transluce looked at dozens of AI models, measuring each across 14 mental-health-related behaviors.
• The organization also worked directly with OpenAI, Anthropic and Google to more deeply understand how people interact with their models.

What they did: Labs scanned a week of their user chats and sent Transluce anonymized patterns — not the conversations themselves — that Transluce used to make its simulated users more realistic.

The big picture: OpenAI, Google and others face multiple lawsuits over instances where people killed themselves after discussing suicide with AI systems, while state and federal regulators have also expressed concerns.

• Other research has also shown the leading chatbots improving at detecting overt indications of suicidal intent, but struggling with subtle cues over longer conversations.

Between the lines: The study found that helpful and harmful behaviors increasingly coexist in the same response.

• A model may tell a user to seek support while still supplying the very material the user sought, such as suicide-related creative writing, practical death preparation or narratives that encourage delusional thinking.
• Older systems were more likely to produce harmful content without any safety-oriented response. But it also highlights that adding a hotline or expression of concern alone doesn't solve the problem.

What they're saying: "There are always going to be failures. These systems are always going to interact with users in surprising ways," Schwettmann said.

• "The way you solve that is not by creating the perfect model, but by being able to kind of anticipate those failures and edge cases in advance," Schwettmann said.

The leading labs said they are working to make their systems better.

• "Google has, for years, helped people find high-quality information and crisis support in the moments they need it most, and we are applying this same research-backed approach to our AI tools," Google senior director Megan Jones Bell said in a statement to Axios.

What we're watching: Transluce plans to open-source its evaluation tools by the end of the year, Schwettmann said.

• The aim is to adapt its approach to other domains, potentially including eating disorders, harmful manipulation and political persuasion.

If you or someone you know needs support now, call or text 988 or chat with someone at 988lifeline.org. En español.

TrustedRouter, an AI routing startup, raised $1.25 million from a group of investors, Axios has learned.

Why it matters: The interest in the startup is the latest example of Silicon Valley's obsession with routing, and the bet that businesses won't rely on just one AI model.

Driving the news: Instead of sending every task to a single AI model, TrustedRouter lets users send queries to the one that fits their needs at the most efficient price.

• The company, which has been in public beta since May, recently processed more than 1 billion tokens in a single day, founder Joseph Perla tells Axios.
• The platform gives users access to more than 600 models from more than 81 providers through a single API, while its own code is entirely open-source.

Follow the money: The seed round includes Sam Lessin of Slow Ventures, early Zoom and Canva investor Bill Tai, 23andMe co-founder Linda Avey and former Stripe data leader George Xing.

• On AI adoption, businesses are "struggling with the double-edged sword of how do I get my employee base up to speed so I don't lose, but how do I not blow my annual budget in six weeks?" Tai said.
• Tai tells Axios that within that environment, the need for routing is "universal."

The big picture: Routing is the shiny object getting the world's attention at the moment.

• Stripe agreed to buy unicorn OpenRouter for more than $8 billion, the company announced earlier this month.
• Meta is reportedly working on an OpenRouter competitor called Switchboard, per The Information.

Zoom in: Perla believes his competitive advantage is a focus on protecting customer data.

• TrustedRouter's pitch is that businesses shouldn't have to choose between model flexibility and data privacy: Its open-source system uses confidential computing.
• That prevents the router from seeing customers' prompts, while letting them choose which AI providers ultimately receive their data if they want that for certain use cases.

What they're saying: Lessin believes routing is an "important" part of the AI story, as it gives customers a "trusted, private, secure way" to use AI tools without "leaking data."

Threat level: Routing threatens to turn AI models from destinations into interchangeable commodities.

• "I see people excited about us and moving to us because they don't like what Anthropic is becoming," Perla said.

AI has an image problem. And one way to fix it is for top companies to dive headlong into health care.

Why it matters: Saving the world with AI-designed cures is better than being blamed for ruining the environment or driving up Americans' utility bills.

• But even the most advanced AI models can't produce miracle treatments at the moment — and could create serious security risks if they try.
• That could set up a fraught debate over the next two or three years about the allocation of resources and whether the benefits of throwing vast amounts of computing capacity into medical research outweigh the costs.

Driving the news: The Wall Street Journal reported this month that Anthropic was trying to shore up investor confidence ahead of its massive initial public offering with talk of pushing harder into health care and biology.

• CEO Dario Amodei wrote on X that Anthropic is moving quickly in biology and medicine "and we hope to have incredible results in the coming years and some early glimmers in the coming months."

Keep reading.

• Music publishers are suing Anthropic, alleging "blatant theft" of copyrighted music. (Axios)

Google has changed the way it refers to Lake Ontario for U.S. users to match President Trump's new moniker, while MapQuest debuted a feature that allows anyone to rename Lake Ontario and share their custom map with others.