All about Artificial intelligence
Despite the enthusiasm around the Artificial Intelligence Agents – Systems capable of making decisions and acting autonomously – The future of this technology is still uncertain.
According to one report From consulting firm Gartner, more than 40% of projects in this area should be discontinued by the end of 2027, pressured by the high implementation cost and lack of clarity on commercial return.
Giants like Salesforce and Oracle bet on high technology, investing billions with the expectation of reducing expenses and increasing efficiency. However, the market begins to give signs of early saturation.
Gartner points out that many suppliers are practicing what they call “agents wash”: they rename simple solutions such as chatbots and virtual assistants, as if they were autonomous agents, when they actually lack really intelligent features.
Read more:
“Today, most AI AGUE projects are still experimental, moved by exaggerated and often poorly applied expectations,” says Anushree Verma, Gartner’s senior director analyst.
“These solutions often do not deliver real value, as current models do not yet have the maturity necessary to deal with complex business goals or adapt continuously to different instructions,” he adds.
Projections for the future
- Despite current challenges, the tendency of the adoption of AI Agentic continues to grow, with positive predictions.
- By 2028, 15% of daily decisions in the workplace should be made autonomously by AI agents – a significant jump over the current 0%;
- In the same period, 33% of corporate software should incorporate AI agent, compared to less than 1% observed in 2024.
For consulting, the potential of technology lies in its ability to go beyond traditional script bots and assistants, offering new forms of complex task automation, resource optimization and innovation in business models.
Strategy is essential
However, Gartner warns: Adopting Age Aggentic without a clear strategy can be a shot in the foot. Integrating intelligent agents with legacy systems can be technically difficult, interrupting already consolidated processes and requiring costly changes.
In many cases, restructuring workflows from the outset, thinking of the logic of Ai Ai, is the most efficient way.
“To extract real value, companies must focus on overall productivity, not just automate isolated tasks,” says Verma.
“It is possible to start with agents to support decision -making, use repetitive routine automation and resort to simple searches assistants. The ultimate goal should be to add value through lower costs, more quality, higher speed and scale.”
