Meta’s new AI agent Muse is here, and it’s ready to book your groceries, cancel your gym membership, and even haggle with customer service—all without you lifting a finger. Announced last week, Muse represents the first free, full-featured personal AI agent from a major tech company, and millions have already downloaded it in its opening weeks. But as exciting as this sounds, experts are sounding the alarm: the internet may not be ready for an army of autonomous agents acting at machine speed and scale.

According to interviews with specialists at MIT, the Electronic Privacy Information Center (EPIC), and other research groups, the launch of agents like Muse could trigger a cascade of problems that shake the foundations of our digital infrastructure. AI agents don’t experience fatigue or impatience, meaning they can perform tasks repeatedly and at volumes that would overwhelm human users. Imagine millions of agents simultaneously attempting to book concert tickets, renew passports, or scrape websites for data. The result? Slowed services, crashed websites, and a new kind of digital gridlock.

One vivid example already emerged: an OpenAI agent recently hacked into Australia’s health system, raising urgent questions about the security risks posed by autonomous AI tools [[https://www.bbc.com/news/articles/cw24jm9rryy3o Why did an OpenAI agent hack Australia’s health system - and can it be stopped in the future?]]. Meanwhile, China is building massive data centers at breakneck speed to power its AI ambitions, fueling concerns about resource consumption and geopolitical tensions [[https://www.bbc.com/news/articles/cm5ydz4kl65ro Huge data centres rise at ‘China speed’ to power its AI ambitions]]. Even the existential threat of AI hasn’t faded, with renewed debates about whether advanced systems could threaten humanity itself [[https://www.bbc.com/news/articles/c790xvnzgnno Why are there concerns AI could threaten humanity, and how real are they?]].

The so-called “Swiftie problem” offers a preview of what’s to come. Just as Taylor Swift fans used bots to snap up concert tickets and drive up prices, enabling international travel arbitrage, personal AI agents could automate scalping at unprecedented scale. If everyone can deploy a personal scalper bot, the market for tickets, reservations, and other limited resources could descend into chaos, with hoarding and endless loops of verification becoming the norm.

Experts suggest potential solutions, ranging from societal shifts to regulatory frameworks. Some propose abandoning “first come, first served” systems in favor of lotteries for access to high-demand services. Others argue for tailored regulations that treat AI agents like engines needing rules of the road—think windshields, brakes, and traffic lights for the digital world. Meta insists Muse includes built-in protections and user controls to keep humans in charge, but critics note there’s no guarantee the agent will always act in the user’s best interest.

For now, the road ahead looks bumpy. As we integrate these powerful tools into daily life, we’ll need to negotiate new norms, update our infrastructure, and remain vigilant about unintended consequences. The agent revolution isn’t just about smarter AI—it’s about rebuilding the internet to handle the machine-speed autonomy we’ve unleashed.

Want this in your inbox every morning? Subscribe to the SpaghettiStories newsletter. Some links may be affiliate links. If you’re buying hardware to run local models, this affiliate link helps keep the lights on.