Become a TechRadar Insider
This week all eyes are on Elon Musk's lawsuit against OpenAI CEO Sam Altman. It's the culmination of a feud that's raged for years, mostly playing out online until now, and it's likely to have huge repercussions for not only the future of OpenAI, but the future of the AI industry as a whole. So far, the courtroom details have already been extraordinary — and not always in ways that flatter.
Alongside the trial, several other big AI news stories caught my attention this week. Lots, as always, involving OpenAI — from news that the company may soon launch a phone to the discovery that a coding model was told not to reference goblins or mythical creatures.
Article continues belowAs always, I’ve pulled together the top stories you need to know below. Think you were paying attention to my round up of the latest AI news from last week? Take the quiz below to find out.
Were you paying attention last week? Take my AI news quiz
The top AI headlines from the past week
Welcome to ICYMI AI, your weekly round-up of the most important developments in artificial intelligence. Here are the biggest AI stories from last week and why they matter.
Musk vs Altman trial is the biggest drama in AI
Elon Musk is suing OpenAI, CEO Sam Altman and the company’s co-founder Greg Brockman for a huge $130 billion. He’s arguing that they betrayed the company's original plan to be a non-profit. However, his motivations are complicated by the fact that he runs his own AI company, xAI. There's more to this trial beyond the drama, though.
At its core, this trial really feels like a test of whether AI companies can ever stick to a mission, especially remaining a non-profit, once money, compute and competition scale up to dizzying levels. If Musk wins, OpenAI could face a very messy restructuring that’ll impact leadership, funding and product development. If he loses, it’ll reinforce that building the latest AI tech may always pull companies toward commercial priorities, no matter how they started out.
Either way, it puts a lot of pressure on external regulation to fill the gap that good intentions alone can't close.
- Read more: ‘This is a real risk, we all could die as a result of artificial intelligence’ — the OpenAI trial took a dramatic turn as Elon Musk and Sam Altman faced off over AI’s real-world danger
- Read more: Letting OpenAI win risks 'losing every charity in America' Elon Musk says in court, as his lawsuit hearing begins
OpenAI might be building an AI-packed phone
Reports emerged this week that OpenAI, the makers of ChatGPT, may be building a smartphone. Analyst Ming-Chi Kuo reported that the device is in development with MediaTek and Qualcomm working on a custom chip and Luxshare handling manufacturing. Speculation is that mass production is targeted for 2028. However, OpenAI hasn't commented.
It looks like the concept behind this OpenAI phone will replace apps with AI agents, maintaining context and completing tasks on your behalf — similar to Google’s plans to turn AI into a layer over everything that we reported on last week.
But think about the broader implications, if OpenAI controls the hardware, it means it’ll bypass big tech players like Apple and Google entirely. As we said last week, how you feel about OpenAI increasingly trying to build an ecosystem rather than just be a chatbot depends on how much you trust OpenAI.
Are young people already sick of AI?
A new report from The Verge that’s certainly ruffled a few feathers this week finds that the more young people use AI, the less they like it. According to the report, despite being among the biggest adopters of chatbot tools, Gen Z workers and students are increasingly resentful of what many describe as an AI-centric future being forced on them — with some actively choosing career paths where they'll never have to use it.
This matters because for the past year, the dominant story has been adoption, especially among young people who I’ve seen described as "AI natives" many times. But this is one of several clear signs of friction in recent weeks. If the people expected to build, use and normalize AI long-term are already losing trust, it complicates the idea that the technology will just slot smoothly into everyday life.
More AI news you might've missed
- China’s power grid will soon be run by an army of humanoid robots: According to the South China Morning Post, the Chinese government plans to put thousands of robots to work on the nation’s infrastructure. It’s interesting to see what large-scale AI deployment may look like and highlights how China is leaning into rapid, state-backed rollout of AI compared to the US.
- OpenAI tells its latest model to stop talking about goblins: In one of the more light-hearted stories of the week, Wired reports that OpenAI really wants its coding agent to never talk about goblins, gremlins or other animals or creatures. It looks like past iterations of the model assumed “bugs” meant mythical creatures and that’s why OpenAI had to spell things out in new instructions.
- We compared ChatGPT Images 2.0 and Google’s Nano Banana 2: We used real-world prompts, but which AI image generator do you think came out on top? You’ll have to click through to find out, but we were fascinated to find out that although both models were similar, one excelled when it came to realism.
- I tried using ChatGPT to follow The 7 Habits of Highly Effective People step by step: TechRadar’s AI editor Graham used ChatGPT to break down a classic self-help system into a structured plan. The results were surprising and prove ChatGPT can really excel as a self development partner.
- OpenAI hit with lawsuits over failure to report school shooter: We saw this one coming last week. According to reports, OpenAI’s moderation tools flagged Jesse Van Rootselaar’s ChatGPT account for discussions of violence. But after much debate, OpenAI didn’t report her to local law enforcement and instead deactivated her account. Now several families are suing OpenAI for not taking action sooner.
- The Pentagon now has a deal with seven AI companies: Companies like OpenAI, Google, Nvidia and several others have now agreed to ‘any lawful use’ of their tech by the US military. Anthropic wasn’t included. This deal raises all sorts of questions about where the boundaries are, and who decides how these systems are used.
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds.
