News

Corporate artificial intelligence bills rise despite falling token prices

Cyprus Mail · 2026-09-13

AI SUMMARY

• What happened: A report by PwC highlights that companies are facing rising costs associated with artificial intelligence (AI) implementation, despite falling token prices, and emphasizes the need for disciplined AI spending to maintain a competitive edge. • Why it matters: Effective management of AI costs is becoming crucial for businesses as the market for AI models becomes more commoditized, with the potential for significant savings that can be reinvested into further AI initiatives. • What to watch next: Companies are encouraged to adopt new operational models focused on cost assessment, system redesign, linking spending to outcomes, and reinvesting savings, which may lead to a shift in how AI spending is approached and managed in the industry.

**Corporate Artificial Intelligence Bills Rise Despite Falling Token Prices**

As artificial intelligence (AI) continues to evolve, companies are grappling with the rising costs associated with its implementation, even as the prices of AI tokens decline. According to a recent report by PwC, a professional services firm, the ability to manage and control AI spending effectively may soon become a key competitive advantage for businesses.

PwC highlights that while cost-control tools for AI have become more widespread and standardized, merely having these tools is no longer sufficient for companies to maintain an edge in the market. Instead, a new operational model that emphasizes disciplined AI spending could enable businesses to generate savings that can be reinvested into further AI initiatives, creating a compounding advantage.

The report cites a case study of a global technology company that adopted this approach, resulting in a significant reduction in operational costs. By implementing stringent cost-control measures, the company was able to decrease the cost of each AI run by 65 to 80 percent, allowing it to execute three to five times more AI processes within the same budget.

Despite the falling prices of tokens, which are the units of measurement for AI operations, PwC notes that overall AI spending is on the rise. The decrease in token prices has led to increased adoption of AI across various processes, which, while potentially beneficial, can also lead to greater complexity and higher overall expenditures. PwC warns that many companies are using AI indiscriminately, often complicating workflows and inadvertently increasing costs.

One of the challenges highlighted by PwC is that organizations frequently consume more tokens than necessary due to a lack of systems capable of identifying areas of waste. The costs associated with AI can be difficult to track, as tokens accumulate across various stages of AI operations, including planning, tool usage, and information retrieval. Additionally, indirect infrastructure costs are often excluded from initial budgets, complicating financial assessments.

The firm emphasizes the importance of model selection in managing costs, noting that the price for one million tokens can vary widely—from just a few cents to as much as $50—depending on the model and tier chosen. PwC warns that opting for the cheapest model is not always the most economical choice, as lower-quality systems may lead to additional work, poor decision-making, or compliance issues. Instead, companies are encouraged to align model tiers with specific tasks, balancing performance needs against cost considerations.

To address these challenges, PwC proposes a new operating model centered around four key disciplines:

1. **Assessing Cost and Value**: Companies should evaluate the cost and value of AI projects before they are developed.

2. **Redesigning Systems**: Organizations need to redesign their systems to minimize waste.

3. **Linking Spending to Outcomes**: It is crucial to connect AI spending to tangible business outcomes.

4. **Reinvesting Savings**: Businesses should reinvest any savings generated from cost-control measures into further AI projects.

Implementing these measures may involve strategies such as reducing unnecessary context in AI workflows, combining multiple tasks into fewer calls, setting spending limits, and routing work to the most cost-effective models capable of completing the tasks.

PwC advocates for building these controls directly into AI systems, incorporating mandatory budget limits, routing rules, workflow thresholds, and audit trails. Human oversight remains essential, with technology designed to flag decisions for review and provide necessary information to align actions with business priorities.

In the case study mentioned, the technology company not only reduced costs but also improved efficiency, cutting average runtime from 12 hours to just four hours while maintaining output quality.

To effectively manage AI spending, PwC recommends that businesses establish a clear understanding of the cost associated with each AI workflow per business outcome. They also suggest placing AI spending on the agenda of chief financial officers and preparing for a shift towards outcome-based pricing models from suppliers.

Furthermore, the firm advises designating finance owners who possess both accountability and technical support to enforce spending discipline, rather than relying solely on dashboards for monitoring.

PwC concludes that the next wave of competitive advantage in AI will not necessarily go to companies that operate the most powerful models. Instead, it will likely favor those that manage their AI operations with greater discipline, as the underlying systems become increasingly standardized across the industry.

Source: Cyprus Mail
RELATED NEWS

More Stories

All News
News

Consumers swap bulky VR headsets for lightweight XR glasses

• What happened: Global shipments of extended reality (XR) headwear fell 38.7% year-on-year to 1.9 million units in the first half of 2026, with a notable shift...

News

Cyprus Business Now: weekly wrap-up

• What happened: Cyprus and Bahrain signed a memorandum of understanding (MoU) on maritime cooperation, covering areas such as maritime safety, training, and in...

News

Revisiting Cyprus’ doctrine of necessity

• What happened: The article discusses the historical context and implications of the doctrine of necessity in Cyprus, particularly its establishment in the lan...

News

Tales from the Coffeeshop: Will Prez use election to destroy another party?

• What happened: Diko party leader ETHNARCH Junior is struggling to manage internal dissent regarding support for President Christodoulides' potential re-e...

News

Cyprus services show mixed performance as euro area output falls

• What happened: Cyprus experienced mixed performance in its services sector, with professional services and transport showing growth, while accommodation and a...

News

Is Christodoulides opening the door to ELAM? The shifting alliances shaping 2028

• What happened: Nikos Christodoulides' recent appointment of Zafeiro Kastanou, an ELAM candidate, to the Cyprus Youth Board has raised questions about a p...