SoftBank is going to great lengths to double down on OpenAI — including taking on significant debt. After completing a $40 billion investment to become one of the ChatGPT maker’s largest backers, the Japanese conglomerate is now seeking a roughly $40 billion loan with a 12-month term, Bloomberg reports.
The financing would be SoftBank’s largest-ever dollar-denominated deal. The AI investment has helped lift profits, but it is also pressuring SoftBank’s credit profile.
OpenAI is delaying its planned “adult mode,” as it seeks to shore up ChatGPT’s core capabilities before the chatbot can generate erotic content.
A source within OpenAI told tech news site Sources that the company will miss its Q1 target for launching the feature:
“We’re pushing out the launch of adult mode so we can focus on work that is a higher priority for more users right now, including gains in intelligence, personality improvements, personalization, and making the experience more proactive.”
The company said it still believes in “treating adults like adults,” but said it wants to get the experience right. OpenAI has been testing user age estimation technology ahead of the planned release.
“We’re pushing out the launch of adult mode so we can focus on work that is a higher priority for more users right now, including gains in intelligence, personality improvements, personalization, and making the experience more proactive.”
The company said it still believes in “treating adults like adults,” but said it wants to get the experience right. OpenAI has been testing user age estimation technology ahead of the planned release.
Anthropic CEO Dario Amodei said in a public post that the company will sue the Pentagon after receiving a letter from the Department of Defense officially designating Anthropic as “a supply chain risk to America’s national security.”
Amodei says that the effect of the unprecedented designation for an American company is more narrow than originally described, and that most of its customers would not be affected.
“With respect to our customers, it plainly applies only to the use of Claude by customers as a direct part of contracts with the Department of War, not all use of Claude by customers who have such contracts.”
Amodei says the company does not “believe this action is legally sound, and we see no choice but to challenge it in court.”
The CEO also apologized for statements he made in a leaked internal memo in which he claimed that the company was targeted because it didn’t show “dictator-style praise” for President Trump.
“With respect to our customers, it plainly applies only to the use of Claude by customers as a direct part of contracts with the Department of War, not all use of Claude by customers who have such contracts.”
Amodei says the company does not “believe this action is legally sound, and we see no choice but to challenge it in court.”
The CEO also apologized for statements he made in a leaked internal memo in which he claimed that the company was targeted because it didn’t show “dictator-style praise” for President Trump.
Feeling the heat from Anthropic’s success with enterprise customers, OpenAI released GPT-5.4, a new model that excels at “professional work.”
OpenAI says the new model has improved capabilities for “professional tasks involving spreadsheets, presentations, and documents. The result is a model that gets complex real work done accurately, effectively, and efficiently — delivering what you asked for with less back and forth.”
The company says the model has advanced computer use skills and supports up to 1 million tokens of context — a measure of the maximum amount of information that can be read and accessed when generating a response, allowing for more complex tasks.
The company says the model has advanced computer use skills and supports up to 1 million tokens of context — a measure of the maximum amount of information that can be read and accessed when generating a response, allowing for more complex tasks.
