What the Answer Does Not See

As AI makes information easier to access, why are the world’s largest companies paying billions of dollars to gain access to certain information?

The costliest piece of information in a loan agreement may not be the clause AI misinterprets. It is the clause it has never seen.

Even the world’s most advanced model cannot know the actual terms of an agreement it cannot access. Its ability to deliver a highly convincing assessment despite not knowing those terms makes the problem even worse.

On October 22, London Stock Exchange Group (LSEG) announced that it had reached an agreement to acquire Octus, a provider of data and analytics on private credit markets, for $3.535 billion in cash.

Octus specializes in researching loan agreements, financial covenants and debt restructurings that are not readily accessible to everyone, and turning them into verifiable, comparable information that can inform decisions.

The company’s 2026 revenue is expected to exceed $300 million.

Here is what makes this interesting: while there is talk of AI making financial research cheaper, one of the world’s largest financial data providers is agreeing to spend billions of dollars to strengthen its ability to access information.

This is not a contradiction. It is an important indication of where the economic value of information is created.

AI can reduce the cost of processing and interpreting existing information. But it cannot obtain a restricted-access document on its own, nor can it independently guarantee a source’s reliability, currency or legal validity.

And this transformation is not confined to financial data.

On October 20, Salesforce signed an agreement to acquire customer engagement platform Braze at an enterprise value of approximately $5.66 billion.

Braze’s value lies less in storing more information about customers than in enabling companies to use their own customer data in real time and act at the right moment.

There is a significant economic difference between noticing a customer moving away from a purchase decision as it happens and explaining why they left in a report prepared three months later.

LSEG is strengthening its access to information; Salesforce is strengthening its ability to use information in time to act.

The broader implication of these two transactions is this: as AI becomes more widespread, competitive advantage may shift from simply producing good analysis to accessing the reliable information that underpins it and turning the findings into business decisions.

This shift also has implications for company valuations.

The size of a company’s database does not, on its own, justify a high valuation. What matters is how readily competitors can obtain that data, how reliable it is, and whether it has been turned into a service that customers would find difficult to give up.

Acquisition prices also cover more than just data: they include customer relationships, technology, expertise, revenues and growth expectations. Buying a strategically valuable asset is therefore not the same as paying a financially sound price for it.

If the expected growth cannot cover integration expenses, financing costs and the acquisition premium paid, an acquisition that appears successful from a technology perspective may still be a poor capital allocation decision.

For companies in Türkiye, the issue is somewhat different.

According to a survey published by TÜİK in October 2026, AI use among enterprises with at least 10 employees rose from 7.5% to 14% in one year. Of the companies using AI, 51% use it for marketing or sales.

Yet as interest in technology grows, financial conditions are not easing at the same pace.

İSO’s September manufacturing PMI stood at 47.9, in contraction territory. On October 22, TCMB cut its policy rate to 36%, highlighting the weakness of domestic demand. Although capacity utilization rose modestly in October, consumer confidence fell to 84.8.

These indicators do not suggest that every sector is in the same position. They do, however, make the question of when technology’s benefits will translate into cash more important, particularly for companies facing uncertain demand and working capital pressures.

For example, an AI application that improves a company’s sales forecasts can create value. But if that same company does not know when its customers will actually pay, its financing needs may grow alongside its sales.

A company’s free cash flow may even deteriorate as its revenues rise.

For many businesses in Türkiye, the first opportunity may therefore lie not in buying a more advanced AI model, but in connecting sales, inventory, receivables, order and financing data to reveal economic realities they cannot currently see.

While global technology companies are buying information they cannot access, some businesses may be losing value because they cannot use the information they already have.

Technology alone will not close that gap. The data must be verified, the business context understood, and the financial impact of the results measured.

Before deciding how much AI to use, management teams need to identify which information gaps are costing the company, and how much. The investment case should rest on the measurable contribution that closing those gaps would make to collections, costs, risk and cash flow.

Because sometimes the costliest information is not the information we pay to acquire, but the information already within the company that we fail to use in time.