VCreaTek

By Sandeep Mehta, Thought Leader in Data & Artificial Intelligence, 07 Oct 2026 

For enterprises, the real skill is knowing where AI can move fast and where it needs more care. 

Artificial Intelligence is moving quickly. Businesses are experimenting with Generative AI (AI that creates new text, images or code in response to a prompt), automating processes, improving decision-making, and finding new ways to work with data. 

For business leaders, the question is no longer simply “Should we use AI?” The more important question is becoming “Where should we use it, how should we use it, and what should we put around it?” 

Not every AI application carries the same level of business risk. Some applications can be tested, improved, and deployed quickly. Others may influence important operational, financial, regulatory, or customer decisions and therefore require stronger validation, security, human oversight, and governance. 

So instead of talking about putting a brake on AI, we think the conversation should be about building better guardrails. 

Enterprise AI Guardrails

AI Is Already Moving into the Enterprise

AI and Generative AI are no longer limited to technology demonstrations or experimental labs. Businesses are increasingly exploring practical applications across functions and industries. 

In supply chains, AI can support demand forecasting, planning, and exception management, such as flagging a late shipment before it causes a stock-out. 

In pharmaceutical and healthcare businesses, it can help people find and use information buried in documents such as standard operating procedures (SOPs), batch records, and regulatory submissions. 

Across manufacturing and Wider enterprise operations, AI can help Spot patterns, process large volumes of information, support employees, and speed decisions  

The opportunity is real, but adoption on its own does not create business value. 

McKinsey’s latest State of AI survey, published in August 2026, shows how wide the gap is. Nearly nine in ten respondents said their organizations use AI regularly in at least one business function. Yet only 37 percent could point to any impact on EBIT (earnings before interest and taxes, a common measure of operating profit), and only about 6 percent said AI accounted for 5 percent or more of their EBIT and had delivered significant value. 

The real value comes when AI is connected to a real business problem, reliable data, the right process, and an outcome that can be measured. 

The Question Is Not How Quickly We Can Deploy AI.

It Is Where, When and Under What Conditions We Should Deploy It. 

Take two examples. In the first, a company uses AI to help employees find information in internal policies and documents. If an answer is incomplete, the employee can check it before acting. 

In the second, an AI system supports a business process where a wrong recommendation could affect product availability, compliance, customer commitments, or an important operational decision. 

Both use AI, but they do not need the same level of testing, approval, monitoring, or human involvement. 

So, the speed of deployment should reflect the importance and risk of the decision the AI supports. 

Regulators have a similar view. The EU AI Act puts stricter obligations on high-risk uses of AI than on low-risk ones. In pharma, the US Food and Drug Administration (FDA) published draft guidance in January 2025 on AI models used to support regulatory decisions about drugs. It says the evidence needed to trust a model should depend on two things. The first is how much the decision relies on the model’s output compared with other evidence, which the FDA calls model to influence. The second is how serious the harm would be if the decision turned out to be wrong, which it calls decision consequence. 

The Global AI Conversation Is Changing

The wider AI industry is having its own version of this debate, about how quickly more capable AI systems should be built and released. 

Over the past few months, frontier AI companies (the handful of labs building the most advanced general-purpose AI models) have been talking openly about the need for stronger evaluations, monitoring and safeguards as AI capabilities advance.  

In August 2026, OpenAI said it had temporarily slowed some of its scaling work while it tightened its security, monitoring, and alignment safeguards. Anthropic published proposed ways to measure how fast AI development is moving inside frontier labs and said it plans to give independent evaluators access to its internal processes. In September, the heads of both companies publicly called on the industry to pace the development of the most capable models.  

Other technology companies, including Meta and Nvidia, have pushed back against a coordinated slowdown or new regulation, arguing that each company should set its own pace, or that market forces are enough, as the Associated Press reported in September. The argument has turned into a wider debate about who should set the pace and how AI safety should be governed.  

A business does not have to take a side in that debate to learn from it. AI capability is now moving fast enough that evaluation, accountability and governance cannot be left until after deployment. 

That debate is mostly about how fast the most capable models should be built. Most enterprises are not building those models. They are deciding how to use them, so the practical question for them is different.  

Do they need to slow AI down, or do they need to get better at deciding where it can move fast? 

Where VCreaTek Sees the Opportunity

At VCreaTek, we see AI as a strong business enabler, as long as it serves a clear purpose. 

We start with the business problem, work out where AI or Generative AI can add real value, and then build the right mix of technology, process, data, and human involvement around it. 

That approach is particularly relevant in areas such as supply chain and pharmaceutical operations, where business decisions can involve multiple systems, large amounts of data, operational dependencies, and, in some cases, regulatory considerations. 

We don’t think every problem needs AI. Where it does fit, it should be understandable, measurable, and aligned with the way the business actually runs. 

Right AI. Right Process. Right Guardrails. Right Outcome. 

Taken together, these four principles give a business a clearer way to move forward. 

Moving Forward

The AI race is not going to stop, and businesses should not wait for it to. But the organizations that do their best in the next phase of enterprise AI may not be the ones that adopt it fastest. 

They are more likely to be the ones that understand where AI creates value, where people must remain involved, and what guardrails allow AI to be deployed with confidence. 

At VCreaTek, our role is to help make AI practical, responsible, and valuable for the businesses that depend on it. 

We believe you can be excited about AI and careful about how it is used at the same time. 

 For us, responsible AI means making AI work better for the business, for its people, and for the outcomes that matter. 

The future of enterprise AI depends on knowing where to accelerate, where to validate, and where human judgment still matters. 

Thinking about where AI can create real value in your business? 

Start with the business problem, not technology. 

Talk to VCreaTek about building the right AI approach for your organization. 

McKinsey & Company “The State of AI in 2026: On the Road to ROI,” August 2026.
View original source
European Union Regulation (EU) 2024/1689—the EU Artificial Intelligence Act.
View original source
US Food and Drug Administration “Considerations for the Use of Artificial Intelligence to Support Regulatory Decision-Making for Drug and Biological Products,” draft guidance, January 2025.
View original source
OpenAI “Pacing Model Development in an Era of Cyber-Critical Capabilities,” 18 August 2026.
View original source
Anthropic “Measurements for Understanding the Pace of AI Development Inside Frontier Labs,” August 2026.
View original source
Axios “Anthropic, OpenAI CEOs Call for Slowdown in AI Development,” 12 September 2026.
View original source
Associated Press “Divisions Emerge in the Tech Industry Over Calls for a Coordinated AI Slowdown,” 16 September 2026.
View original source

Sources

  1. McKinsey & Company, “The State of AI in 2026: On the Road to ROI,” August 2026. Read the McKinsey report
  2. European Union, Regulation (EU) 2024/1689—the EU AI Act. Read the EU AI Act
  3. US Food and Drug Administration, “Considerations for the Use of Artificial Intelligence to Support Regulatory Decision-Making for Drug and Biological Products,” January 2025. Read the FDA guidance
  4. OpenAI, “Pacing Model Development in an Era of Cyber-Critical Capabilities,” 18 August 2026. Read the OpenAI publication
  5. Anthropic, “Measurements for Understanding the Pace of AI Development Inside Frontier Labs,” August 2026. Read the Anthropic publication
  6. Axios, “Anthropic, OpenAI CEOs Call for Slowdown in AI Development,” 12 September 2026. Read the Axios article
  7. Associated Press, “Divisions Emerge in the Tech Industry Over Calls for a Coordinated AI Slowdown,” 16 September 2026. Read the Associated Press report
Leave a Reply

Your email address will not be published. Required fields are marked *

Translate »