5 generative AI trends to watch in 2025

Generative AI is as trendy as ever.

This year, AI research was awarded Nobel Prizes, and the world’s largest technology companies pumped AI into as many products as possible. The US government promoted AI as an engine for creating a clean energy economy and as a strategic pillar for federal spending. But what’s next for 2025?

The trend of generative AI in the last few months of 2024 points to a greater push for adoption by tech companies. Meanwhile, results on whether AI products and processes deliver ROI for enterprise software buyers are mixed. While it is difficult to predict how AI will continue to shape the technology industry, experts have made predictions based on current trends.

Respondents to a IEEE study rated AI in September as one of the top three technology areas that will be most critical in 58% of cases by 2025. Conversely, almost all respondents (91%) agree that by 2025 there will be “a generative AI reckoning” regarding what the technology can or should do. Expectations for generative AI are high, but the success of projects that use it remains uncertain.

1. AI agents will be the next buzzword

Based on my research and observations, the use of AI tools will increase significantly by 2025.

AI agents are semi-autonomous generative AI that can be linked together or communicate with applications to execute instructions in an unstructured environment. For example, Salesforce uses AI agents to calling sales leads. As with generative AI, the definition of an agent’s capabilities is unclear. IBM defines it as an AI that can reason through complex problems, such as Open AI o1. However, not all products billed as AI agents can reason that way.

Regardless of their capabilities, AI agents and their use cases will likely be at the forefront of generative AI marketing by 2025. AI ‘agents’ would be the next phase of evolution for this year’s AI ‘copilots’. AI agents could spend time independently completing multi-stage tasks while their human counterpart handles another task.

2. AI will both help and hurt security teams

Cybersecurity attackers and defenders alike will continue to benefit from AI in 2025. By 2024, there will already be an increase in generative AI security products. These products can write code, detect threats, answer tricky questions or serve as a ‘rubber duck’ during brainstorming sessions.

But generative AI can present information that is inaccurate. Security professionals can spend as much time double-checking the output as if they had done the work themselves. Failure to review such information may result in broken code and even more safety problems.

“As AI tools like ChatGPT and Google Gemini become deeply integrated into business operations, the risk of accidental data exposure increases with new data privacy challenges,” said Jeremy Fuchs, cybersecurity evangelist at Check Point Software Technologies, in an e-mail email TechRepublic. “By 2025, organizations must take swift action to implement strict controls and governance over the use of AI to ensure that the benefits of these technologies do not come at the expense of data privacy and security.”

Generative AI models, like any other software, are susceptible to malicious actors, especially via jailbreak attacks.

“The growing role of AI in cybercrime is undeniable,” Fuchs explains. “By 2025, AI will not only increase the scale of attacks, but also their sophistication. Phishing attacks will be harder to detect because AI is constantly learning and adapting.”

Generative AI can make conventional methods of identifying phishing emails (poor grammar or out-of-the-blue messages) obsolete. Securing disinformation will become more important as AI-generated videos, audio, and text spread. As a result, security teams must adapt to both usage and security defending against generative AI – just as they have adapted to other major changes in business technology, such as the large-scale migration to the cloud.

3. Companies will evaluate whether AI delivers ROI

“The pendulum has swung from ‘new AI innovation at any cost’ to a resounding imperative to prove ROI in boardrooms around the world,” said Uzi Dvir, global CIO at digital adoption platform company WalkMe, in an email. “Similarly, employees are wondering whether it is worth the time and effort to figure out how to use these new technologies for their specific roles.”

Organizations have difficulty determining whether generative AI adds value and for which use cases it can make the most difference. Organizations that adopt AI often face challenges high costs and unclear objectives. It can be difficult to quantify the benefits of generative AI use, where these benefits manifest themselves and what they compare to.

This challenge is a side effect of the integration of generative AI into many other applications. It makes some decision makers wonder whether generative AI add-ons really increase the value of those applications. Levels of AI can be costly, and in the coming year it’s expected that more companies will rigorously test (and sometimes discard) the features that aren’t delivering results.

Many companies that integrate generative AI at scale are seeing success. Be with Q3 earnings callGoogle attributed this result to his AI infrastructure and products such as AI overviews. However, Meta reported that AI could increase significantly capital expenditureeven if the number of users decreases.

SEE: Google Cloud previews the sixth generation of the AI accelerator Trillium.

4. AI will have a major impact on scientific research

Today’s AI is not only impacting enterprise productivity, but has also seen significant changes in science.

Four of the 2024 Nobel Laureates used AI:

• Demis Hassabis and John Jumper from Google DeepMind won the Nobel Prize in Chemistry for predicting the structure of proteins with AlphaFold2.
• John J. Hopfield and Geoffrey Hinton won the Nobel Prize in Physics for their decades of work on the development of neural networks.

The White House held a summit on October 31 and November 1 the use of AI in the life scienceshighlighting how AI enables solutions to complex challenges in ways that impact the world. This trend will likely continue next year as generative AI models grow and mature.

5. The environmental tools created with AI will not offset the energy toll

Energy efficiency is another buzzword in AI.

But for every use case where AI can help predict weather patterns or optimize energy use, there’s a different story about the environmental costs of building the data centers needed to perform generative AI. Such construction requires enormous amounts of electricity and water – and rising global temperatures are only exacerbating the problem. It is unlikely that equilibrium will be reached in this large-scale problem.

However, expect companies to push dubious and genuine claims of energy savings and environmental friendliness in AI. Consider resource usage related to your organization’s AI strategy.

What are the most popular generative AI products?

The most well-known generative AI products are:

• ChatGPTthe OpenAI chatbot
• Googling Twin
• Microsoft Second pilot
• GPT-4the major language model behind ChatGPT
• DALL-E 3, an image generator

What is the most advanced generative AI?

Several tests have been proposed as potential criteria for determining the most advanced generative AI. Some organizations judge their models against human educational standards, such as the International Mathematical Olympiad or Codeforce competitions.

Other assessments, such as measuring Massive Multitask Language Understanding, are made explicitly for generative AI. Google’s Gemini Ultra, China Mobile’s Jiutian and OpenAI’s GPT-4o top the list. MMLU Ranking Today.