Gartner: 2025 will see the rise of AI agents (and other top trends)

The pace of AI continues to accelerate, with possibilities never previously thought possible now becoming a reality. This is especially true for AI agents, or virtual colleagues, who will work alongside us and ultimately autonomously.

Gartner predicts that at least by 2028 15% of daily work decisions will be made autonomous via agentic AI ​​(compared to 0% in 2024). To further highlight the technology’s potential, the company has named it one of its top strategic technology trends for 2025.

“It’s happening very, very quickly,” Gene Alvarez, distinguished VP analyst at Gartner, told VentureBeat. “No one ever goes to bed at night when everything is done. Organizations spend a lot of time monitoring issues. The ability to create agents that not only do that monitoring, but also take action will help not only from a productivity perspective, but also from a timing perspective.”

What else does Gartner predict for the coming year? Here are some trends the company will explore throughout its operations Gartner IT Symposium/Xpo 2024 this week.

AI agents both ‘cool and scary’

The entry-level model for AI agents consists of mundane tasks that consume human time and energy, Alvarez explains.

The next level is agentic AI that can monitor and manage systems autonomously. “Agentic AI has the ability to plan, sense and take action,” says Alvarez. “Instead of having something that just monitors systems, agent AI can do the analysis, create the solution and report back that it happened.”

Looking at even more complex scenarios, agents could one day help scale workforces. For example, a new employee who would normally shadow a human can be mentored by an AI colleague.

“You can have an agent as a mentor so they can climb the learning curve much faster,” says Alvarez.

He acknowledged that all of this is “cool and scary” at the same time, and that there is fear of job loss. “But if the agent can actually teach me a new set of skills, I can move from a job that is going away to a job that is needed,” he emphasized.

Systematically building trust in AI

Moving on to the next top trend, Alvarez noted, “There is a whole new workforce, how do we manage that?”

This will give rise to AI governance platforms, which enable organizations to manage the legal, ethical and operational performance of their AI systems. New tools will create, manage and enforce policies to ensure AI is transparent and used responsibly. These platforms can check for bias and provide information about how models are built, as well as the reasoning behind their cues.

Ultimately, Alvarez predicted, such tools will become part of the AI ​​creation process itself to ensure ethics and governance are built into models from the start.

“We can create trust through transparency,” he said. “If people lose confidence in AI, they won’t use it.”

Not just one type of computer model

There are seven computing paradigms “right on our doorstep right now,” Alvarez pointed out. These include CPUs, GPUs, edge, application-specific integrated circuits, neuromorphic systems, classical quantum and optical computers.

“We always had the idea of ​​going from one to the other,” Alvarez says. “But we were never quite successful in completing that step.”

But the hybrid computing models of the future will combine different computing, storage and networking mechanisms, he noted. Orchestration software moves computing power from one to another depending on the task and the method best suited to the task.

“It’s about how we can get them to work together,” Alvarez said.

At the same time, new, more specific computing technologies will consume significantly less energy, he pointed out. This is important as there is increasing pressure to reduce consumption and carbon footprint. But “at the same time, demand for IT computing capabilities is increasing incredibly rapidly.”

Incremental improvements will not be enough; Companies need long-term solutions, he said. New technologies – such as green cloud providers or new, more efficient algorithms – could improve efficiency by thousands or even tens or hundreds of thousands of orders of magnitude.

Proactively addressing disinformation security

AI allows threat actors to spread disinformation faster – and easier – than ever before. They can push out deepfakes and craft convincing phishing emails; exploiting vulnerabilities in workforce collaboration tools; use malware to steal login credentials; and initiating account takeovers (among other tactics).

This makes the security of disinformation crucial; the emerging category seeks to assess authenticity, track the spread of malicious information, and prevent impersonation. Elements include brand impersonation scanning, third-party content evaluation, claim and identity verification, phishing mitigation, account takeover prevention, social/mass media and dark web monitoring, and sentiment manipulation. Deepfake detection will also be able to identify synthetic media, Alvarez explains, and watermarking tools will ensure users are interacting with real people.

Gartner predicts that in 2028 half of all companies will adopt products, services or features specifically designed for disinformation security, compared to less than 5% today.

“Disinformation security won’t just be a single technology,” Alvarez said, “it will be a collection of technologies.”

Preparing security for the post-quantum world

Currently, the Internet runs on public key cryptography, or asymmetric encryption, which secures two points of communication. This encryption is difficult to break because it simply takes too long, Alvarez explains.

However, quantum is developing rapidly. “There will come a point where quantum computing will work and we will be able to break that encryption because it has the mathematical power to do that in real time,” Alvarez says.

Red teams are already gearing up and waiting: many are collecting encrypted data and holding it until quantum is realized. That won’t last long: that’s what Gartner predicts by 2029Advances in quantum computing will make most conventional asymmetric cryptography insecure.

“We believe it will be bigger than the millennium, if not bigger,” Alvarez said.

Organizations must prepare for post-quantum cryptography now, he said, to ensure their data can withstand decryption. Alvarez pointed out that it is not easy to switch cryptography methods and that it is “not a simple patch.”

A good place to start has been established standards from the National Institute for Standards and Technology (NIST). Alvarez pointed out that the agency will release the second version of its guidelines for post-quantum cryptography in the spring of 2025.

“What do you do when all the locks are broken? You need new locks,” Alvarez said. “We want to make sure we update our security before quantum becomes a reality.”

AI improves our brains

Going deeper into the sci-fi arena, Gartner expects an increase in the use of bidirectional brain-machine interfaces (BBMIs) that read and decode brain activity and enhance human cognitive skills. These can be integrated directly into our brains or enabled through wearables such as glasses or headbands, Alvarez explains.

Gartner expects that by 2030, 30% of knowledge workers will use technologies like BBMIs to stay relevant in the AI-powered workplace (up from less than 1% in 2024). Alvarez said he sees potential in upskilling people and next-generation marketing – for example, brands will be able to know what consumers are thinking and feeling to gauge sentiment.

Alvarez ultimately compared it to the 2011 film ‘Limitless’ or Apple TV’s ‘Severance’ (although, to be fair, neither portrays the technology in the most positive light). “It can reach your brain and improve its function,” he said.