July 26, 2025
5 min read
PYMNTS
Learn to identify ‘agent washing’—vendors mislabeling AI tools as autonomous agents—and avoid costly AI investments without real agentic capabilities.
Companies of all sizes are being deluged by vendors promoting the latest artificial intelligence (AI) trend: AI agents. However, most AI agent systems sold today are not truly agentic, according to a Gartner report.
Out of thousands of AI agent systems marketed, only about 130 are genuinely agentic. Agentic AI refers to systems with autonomy to plan, reason, and act toward goals with limited or no human input. Unfortunately, many vendors conflate these with simpler tools lacking such capabilities.
“Many vendors are contributing to the hype by engaging in ‘agent washing’ — the rebranding of existing products, such as AI assistants, robotic process automation (RPA), and chatbots, without substantial agentic capabilities,” Gartner noted.Vendor selection is critical. While CFOs have grown more comfortable with generative AI, many remain uncertain if agentic AI is ready for enterprise deployment, according to recent PYMNTS Intelligence reports. Gartner predicts over 40% of agentic AI projects will be canceled by 2027 due to high costs, unclear business value, and weak risk controls stemming from AI systems incorrectly marketed as agentic.
What Defines a True AI Agent?
Sagi Eliyahu, co-founder and CEO of Tonkean, explains that true AI agents exhibit goal-driven autonomy — they work dynamically and proactively with self-determination to pursue long-term business objectives. These agents can use tools, leverage unique skills, and collaborate with other agents to complete complex tasks across technology environments. However, integration is key. Eliyahu emphasizes:“If the ‘agent’ only handles discrete tasks defined by the user, works only within its own system, or is accessible only through chat, it’s not an agent — it’s automation or a chatbot.”Akhil Sahai, Chief Product Officer at Kanverse.ai, suggests companies ask these questions to identify true agentic AI:
- Can the system operate without constant human input?
- Does it pursue goals autonomously rather than follow scripted tasks?
- Can it reason, plan, and improve with experience? If the answer to any is “no,” it’s not an AI agent.
- Agent washing is widespread, with vendors mislabeling basic automation or LLM apps as AI agents, causing confusion and wasted investments.
- True AI agents autonomously collaborate to achieve shared goals, reason, plan, and adapt.
- Business leaders should demand proof and ask critical questions before investing.
- The Two Faces of AI: Gen AI’s Triumph Meets Agentic AI’s Caution
- AWS Unveils AI Agent Marketplace as ‘One-Stop Shop’ for Enterprise Deployment
- Meet the AI Agents Acing Compliance Tests and Correcting Government Data
- AI Agents Do Well in Simulations, Falter in Real-World Shopkeeping Test
- India Crypto News: India Deploys AI to Track Crypto Tax Evaders
- Streamline Financial Services Operations with AI-powered Agents
- AI Crypto Coins Drive 2025 Innovation as Blockchain and AI Converge
Demand for Hard Evidence
Eliyahu stresses the importance of orchestration — multiple bots coordinating tasks to achieve a shared goal with minimal human intervention. Unlike traditional automation that performs fixed sequential tasks, agentic AI can plan, adapt, and collaborate.“Agentic orchestration is how you instrument AI agents for enterprise. It puts agents alongside employees to coordinate workflows, execute tasks, and drive outcomes — all while following configurable policies and guardrails.”However, the term “AI agent” has become a marketing buzzword. Sahai warns:
“The term has become a catch-all, slapped onto everything from basic workflow automation to applications that simply call a large language model (LLM) for a response. This isn’t harmless marketing; it fuels confusion about what AI agents truly are.”This trend resembles earlier “cloud washing” and “AI washing,” where legacy software was relabeled to appear modern. Simple rule-based automation and apps requiring frequent human input are often mislabeled as agentic AI. These exaggerated claims risk eroding trust, causing confusion, stagnation, and wasted investment.
“Most agentic AI propositions lack significant value or return on investment (ROI),” said Anushee Verma, senior director analyst at Gartner.Despite current challenges, Gartner sees long-term potential: by 2028, 15% of daily work decisions will be autonomously executed, and 33% of enterprise software applications will include agentic capabilities — up from less than 1% today.
Buyer Beware
Sahai advises business leaders:“Don’t settle for AI agent as a label. Demand evidence. Ask hard questions. Avoid repeating the cycle we saw with cloud-washing and AI-washing.”
Key Takeaways:
For more insights on AI and digital transformation, subscribe to the PYMNTS AI Newsletter. Read more:
For the original article, visit PYMNTS.