The era of the All-Access AI agent has arrived

For years the cost of using “free” services from Google, Facebook, Microsoft and other Massive Tech companies transfer your data. Uploading your life to the cloud and using free technology provides convenience, but it puts your personal data in the hands of giant corporations that will often want to profit from it. Now the next wave of generative AI systems will likely need more access to data than ever before.

Over the past two years, generative AI tools — such as OpenAI’s ChatGPT and Google’s Gemini — have moved beyond the relatively plain, text-based chatbots that companies initially released. Instead, Massive AI is increasingly evolving into agents and “assistants” that promise to be able to take actions and perform tasks on your behalf. Problem? To take full advantage of their capabilities, you must grant them access to your systems and data. While much of the initial controversy surrounding vast language models (LLMs) concerned the blatant copying of copyrighted data on the Internet, access by AI agents to your personal data will likely create a number of modern problems.

“To be fully functional, to be able to access applications, AI agents often need to gain access to the operating system or operating system level of the device they are running them on,” says Harry Farmer, a senior research fellow at the Ada Lovelace Institute, whose work has included research on, among other things, the impact of AI assistants and found they could pose “deep threats” to cybersecurity and privacy. Farmer says there may be data trade-offs when personalizing chatbots or assistants. “All these things need a lot of information about you to work,” he says.

While there is no strict definition of what an AI agent actually is, it is often best to think of it as a generative AI or LLM system that has been given some level of autonomy. Right now, agents or assistants, including AI web browsers, can take control of your device and browse the Internet, book flights, conduct research, or add products to your cart. Some people can perform tasks consisting of dozens of individual steps.

Although current AI agents have problems and often cannot complete tasks they decided to do ittech companies are betting that the systems will essentially do just that Change jobs for millions of people as their competences augment. A key part of their usefulness likely comes from access to data. So if you need a system that can keep you on track with your schedule and tasks, it will need access to your calendar, messages, emails, and more.

Some of the more advanced AI products and features provide insight into how much access agents and systems can be granted. Some agents developed for companies can read code, emails, databasesSlack messages, files stored in Google Drive, and more. Microsoft’s controversial product Recall takes screenshots of your desktop every few seconds, so you can search through everything you’ve done on your device. Tinder has created an artificial intelligence feature that can search photos on your phone “to better understand” your users’ interests and personality.

Carissa Véliz, author and associate professor at the University of Oxford, says that most of the time, consumers have no real way to know whether AI or technology companies are treating their data the way they claim. “These companies are very promiscuous with their access to data,” Véliz says. “They have shown they have no respect for privacy.”