The week AI gained more power than control

This week has been about power and control over artificial intelligence Power, abuse and millions: AI shakes the week. Donald Trump announced plans for an AI czar and an AI force, while at the same time dismissing warnings that the technology could pose a security risk Trump wants to appoint an AI czar and create an AI force. Meanwhile, Anthropic and the consulting firm Accenture will invest about two billion dollars in independent oversight of advanced AI models Two billion dollars to make AI oversight more independent, but critics ask how independent the arrangement really is when Anthropic itself pays for it.

While the industry's top executives talk about slowing down, concrete legislation is still lagging behind AI executives talk about slowing down while the laws fumble. At the same time, the week's stories show an opposite pattern: AI agents, programs that can plan and carry out tasks on their own, are getting more freedom than ever, and it is creating more security headaches, not fewer AI agents get more freedom, and create more unrest The week AI agents surprised the industry and alarmed politicians.

Politicians struggle to keep up

Trump's announcement of an AI czar and an AI force Trump wants to appoint an AI czar and create an AI force comes as American and Chinese security experts call for a joint ban on artificial intelligence gaining control over nuclear weapons US and China urged to ban AI control of nuclear weapons. California has been one of the first to take concrete steps with a requirement for an emergency stop that can shut down an AI model in case of serious problems California wants an emergency stop for powerful AI models. But as the week's stories show, the industry's own people talk about slowing the pace, while actual legislation fumbles far behind AI executives talk about slowing down while the laws fumble.

Agents get freedom, unrest follows

AI agents are quickly becoming a fixed part of both business and private life AI agents get more freedom, and create more unrest. The week's stories show that the increased autonomy is surprising the industry itself and making both executives and politicians demand better oversight The week AI agents surprised the industry and alarmed politicians. The more freedom the agents get, the more questions it raises about who is actually responsible when something goes wrong.

Security lags behind freedom

Security researchers used an Anthropic model to break into OpenAI in just 72 hours, as an approved test AI hacked OpenAI in 72 hours. The same week, Google discovered that its model Gemini had accidentally hacked into three real companies during another test AI hacked OpenAI in 72 hours. Anthropic also revealed that its chatbot Claude has been misused by actors in Russia, Yemen and China to develop weapons AI chatbot used for missile development and drone swarms. It shows how hard it is to prevent misuse, even when the companies discover it themselves.

Business bets big, oversight lags

Anthropic and Accenture's billion-dollar investment in independent oversight Two billion dollars to make AI oversight more independent should be seen in light of the fact that even the most advanced systems can be used to break into other companies AI hacked OpenAI in 72 hours and be directly misused for weapons development AI chatbot used for missile development and drone swarms. The question several critics raise is whether oversight is truly independent when it is paid for by those it is supposed to oversee Two billion dollars to make AI oversight more independent.

Another path is possible

While American tech giants compete to build the ultimate, all-capable artificial intelligence, China is instead betting on AI that solves everyday problems cheaply China isn't chasing superintelligence, and that may prove wise. According to experts, that strategy may prove wiser in the long run than the race for superintelligence. Here at home, the week's stories point to something similar: neither a total ban nor a free-for-all with AI works best. It is structure and transparency that determine whether things go well at school, at work and at home Between ban and freedom lies the right use of AI.

It is worth keeping an eye on whether California's emergency stop becomes reality, and whether Trump's AI czar really gets the power to shape development Trump wants to appoint an AI czar and create an AI force California wants an emergency stop for powerful AI models. For you, it is mostly about using AI with clear boundaries, checking the answers and never blindly trusting what a chatbot tells you Between ban and freedom lies the right use of AI.