ZDNET’s key takeaways
- Gartner has released its 2025 Hype Cycle report.
- AI agents and data are at their most inflated and need precise application to yield results.
- The report also emphasized trust and safety efforts as critical to the next five years.
Research firm Gartner has released its annual Hype Cycle report, which investigates whether new technology is living up to expectations or is still far off from making a meaningful impact. At the top of the list this year? AI agents.
The report comes at a time when tech companies are making big promises about AI’s capabilities and the extent to which it will become entrenched in our everyday lives, both at home and work.
Gartner named four main technologies: agents, AI-ready data, multimodal AI, and what it called “AI trust, risk and security management (TRiSM).” Let’s break down what it says about each one.
AI agents and AI-ready data
AI agents refer to increasingly autonomous systems that can carry out tasks for humans, ideally with little to no intervention. However, at their simplest, they can function like standard chatbots or assistants; “agent” remains a pretty widely defined word, meaning what agents actually do varies significantly in terms of sophistication and application.
Also: AI agents will be ambient, but not autonomous – what that means for us
AI is only as good as the data it has access to. AI-ready data is information that is correctly structured to be ingested by AI tools, meaning it’s been optimized for efficiency and accuracy. Unlike the other three technologies Gartner’s report focused on, the firm said AI-ready data has a slightly longer plateau timeline of five to 10 years, compared with the two to five-year timeline of the others. The graph’s Plateau of Productivity refers to a point at which a technology proves its practical use and market appeal.
Gartner acknowledged that these two technologies are among the most rapidly advancing, but that they are “accompanied by ambitious projections and speculative promises, placing them at the Peak of Inflated Expectations,” or a critical saturation point where their productive applications are either realized or not.
As AI investments have stayed consistent this year, businesses are looking to AI for “operational scalability and real-time intelligence,” said Haritha Khandabattu, senior director analyst at Gartner. “This has led to a gradual pivot from generative AI (GenAI) as a central focus, toward the foundational enablers that support sustainable AI delivery.”
Also: 5 ways to successfully integrate AI agents into your workplace
With that comes a need for more concrete outcomes to match big promises. As with most applications of AI, Gartner noted that in order to use agents and data effectively, businesses have to consider where it can actually help them strategically, rather than applying them blindly across organizations.
“To reap the benefits of AI agents, organizations need to determine the most relevant business contexts and use cases, which is challenging given no AI agent is the same and every situation is different,” Khandabattu said in the report. “Although AI agents will continue to become more powerful, they can’t be used in every case, so use will largely depend on the requirements of the situation at hand.”
As for data, Gartner said it’s only as effective as a company’s approach to data management. Simply having the data isn’t enough — the report encouraged organizations not to underestimate how much they may need to evolve their data management to meet the requirements of AI.
“This will cater to existing and upcoming business demands, ensure trust, avoid risk and compliance issues, preserve intellectual property and reduce bias and hallucinations,” the report said.
Multimodal AI and TRiSM
Gartner said multimodal AI and risk management infrastructure, despite also “dominating the Peak of Inflated Expectations,” each show promise with fewer caveats. Both will be adopted by the mainstream in the next five years and “will enable more robust, innovative and responsible AI applications, transforming how businesses and organizations operate.”
Multimodal AI is trained on and can output more than one type of data, including audio, text, images, video, and more. This creates more context than text models alone, which Gartner said will open up new applications of AI.
Also: 5 ways to successfully integrate AI agents into your workplace
On the safety front, Gartner’s report said AI TRiSM is central to using AI in secure and ethical ways. AI’s rapid evolution is surfacing new security issues and ethics questions just as quickly; to move AI applications forward properly, “AI brings new trust, risk and security management challenges that conventional controls don’t address,” said Khandabattu. “Organizations must evaluate and implement layered AI TRiSM technology to continuously support and enforce policies across all AI entities in use.”
The Trough of Disillusionment
This year’s report noted that several areas of AI have slipped into the Trough of Disillusionment, or the period in which a technology fails to reach the lofty expectations set for it during the inflation stage. The report puts synthetic data and generative in this section of the Hype Cycle graph, stating that each will plateau in about two to five years.
