FuturismTechno flags agentic AI shift

Futurism Technologies used a May 22 blog post to frame agentic AI as the next stage of procurement software: systems that do not just surface risks or recommend suppliers, but carry out multistep work across sourcing, approvals and compliance. The post, which circulated in social media discussions over the weekend, said procurement teams have invested heavily in dashboards, analytics and sourcing tools but still face slow cycle times and fragmented workflows. (mckinsey.com) McKinsey made a similar argument in a February 5 article, describing a move from analytical AI — “Show me the data” — to agentic AI — “Do it for me.” The firm said AI agents can emulate human judgment, carry out multistep tasks and continuously improve through learning loops. (cloud.google.com) The thread running through both accounts is straightforward: procurement is becoming a test case for software that can analyze, decide and act inside enterprise systems, with humans moved to exceptions, approvals and oversight. That is the shift people were reacting to in the FuturismTechno post. (futurismtechnologies.com) ### What makes “agentic” AI different from the last wave of procurement tools? McKinsey said older procurement technology focused on transactional automation such as purchase orders, invoices, catalogs and sourcing events, while decision-making remained manual and backward-looking. (mckinsey.com) Agentic AI, by contrast, is described as software that ingests complex data, reasons through trade-offs and autonomously generates options and recommendations. Futurism Technologies said the practical difference is execution. Its May 22 post said most AI procurement software still behaves like a “smart advisor,” while agentic systems are built to understand business context, apply rules and compliance automatically, execute procurement steps on their own and involve humans only when judgment matters. (futurismtechnologies.com) ### Where are companies saying these systems can actually be used? Google Cloud said in an October 2025 manufacturing report that 56% of manufacturing executives surveyed said their organizations were actively using AI agents, and 37% said they had launched more than 10. (mckinsey.com) The same survey said adoption was already showing up in quality control, production planning, supply chain and logistics, and factory operations. Bain said AI agents can monitor demand forecasts, supplier performance, market shifts and supply chain risks in real time, then generate negotiation strategies, draft contracts and optimize category decisions. (futurismtechnologies.com) That is close to the use case described in the social posts: sourcing, negotiation and operating decisions handled end to end by software agents. ### How solid is the “40% efficiency gain” claim? McKinsey said the shift to a hybrid workforce of procurement professionals and digital coworkers could make the procurement function 25% to 40% more efficient. Futurism Technologies cited that estimate directly in its May 22 post. (cloud.google.com) IBM’s Institute for Business Value used a different dataset but pointed in the same direction. IBM said organizations using AI in procurement expect a 20% productivity gain, a 14% operational-efficiency increase and, by 2027, 41% greater efficiency in source-to-pay processes. (bain.com) Those numbers are projections or survey findings, not audited results across the whole market. But they help explain why procurement has become one of the most active corporate test beds for agentic systems. (mckinsey.com) ### Why does governance keep coming up in these discussions? Bain said AI cannot compensate for weak governance, fragmented data or broken processes, even as procurement moves toward networks of agentic systems that can initiate actions and execute decisions. Google Cloud also said multi-agent systems act on behalf of users “under supervision.” (ibm.com) That is the operational issue underneath the hype. If software can choose suppliers, prepare negotiation strategies, trigger workflows or draft contracts, companies need rules for authority, audit trails, policy limits and escalation. The near-term question is not whether agents can do more work. (mckinsey.com) It is how much authority companies are willing to delegate, and where they still require a human signoff. That inference is based on the governance warnings in Bain, the supervision language in Google Cloud, and the real-time compliance claims in Futurism Technologies. ### What should readers watch next? IBM said chief procurement officers expect 41% greater efficiency in source-to-pay processes by 2027, while Google Cloud said 55% of manufacturing executives plan to allocate at least half of future AI budgets to AI agents. (bain.com) Those two markers — budget allocation and source-to-pay deployment — are the clearest signals of whether the current discussion turns into broader adoption. (ibm.com)