Turkey Artificial Intelligence Initiative Examines Companies' Approach to Artificial Intelligence

The Artificial Intelligence Research released by Turkey's Artificial Intelligence Initiative (TRAI) has revealed the maturity level of institutions in Turkey regarding artificial intelligence applications. According to the report prepared with the participation of 126 institutions, the share of companies actively using artificial intelligence reached 15.2%, while the share of institutions running five or more projects reached 32.8%.

While 37.6% of companies define their strategy at board level, the areas where institutions face the most difficulty emerged as talent (58%), budget (57%) and organizational culture (55%). The most commonly used models in generative artificial intelligence were OpenAI (80%), Microsoft Copilot (44%) and Google Gemini (42%).

TRAI announces results at Turkey's Artificial Intelligence Summit
TRAI shared the research findings with the public at Turkey's Artificial Intelligence Summit, organized under the theme "Beyond Intelligence". The study comprehensively analyzed institutions' artificial intelligence strategies in Turkey, application areas and major obstacles they face.

According to the data, one out of every six companies in Turkey actively uses artificial intelligence. The fact that the share of institutions running five or more artificial intelligence projects reached 32.8% demonstrates that artificial intelligence is increasingly moving to a strategic position in organizational structures.

One-third of companies define their strategy at board level
The research showed that 37.6% of institutions have approved their artificial intelligence strategy at board level. However, it is noted that a significant portion is still in preparation or draft stages.

The objectives most prominently cited in artificial intelligence institutional strategy are:

-Increasing operational efficiency

-Reducing costs

-Strengthening decision support mechanisms

Furthermore, the number of companies integrating artificial intelligence into their organizational structure has increased, and the share of institutions with dedicated artificial intelligence teams has risen to 15.2%. 42% of these teams consist of 4-10 people and work largely integrated with data analytics units.

Model diversity growing in generative artificial intelligence
The TRAI report shows significant momentum in Turkey's generative artificial intelligence (GenAI) field.

Institutions are widely testing generative artificial intelligence applications in content creation, customer communication, corporate training, coding support and reporting processes.

Model usage rates are as follows:

-OpenAI: 80%

-Microsoft Copilot: 44%

-Google Gemini: 42%

-Anthropic Claude, Meta Llama and Mistral: Rising alternative models

This table demonstrates that institutions are moving toward hybrid model strategies rather than focusing on a single model or ecosystem.

The business areas where generative artificial intelligence is most commonly used are listed in the report as follows:

-Chatbots and corporate assistants

-Copilots and coding support systems

-Content creation and summarization tools

-RAG-based solutions

-Corporate knowledge management systems

Agent-based artificial intelligence applications herald a new era
Agent-based artificial intelligence systems, defined in the research as "the next wave", are beginning to enter the agenda of Turkish companies.

More than half of institutions stated they are conducting pilot projects or proof-of-concept (PoC) studies related to agent systems.
Notable platforms include:

-Microsoft Copilot Studio

-LangChain

-Google AI Agent Builder

In application areas, solutions such as HR Agent, Finance Agent, CRM Agent and Document Assistant are drawing attention.

These findings demonstrate that Turkish companies are transitioning from generative artificial intelligence to "thinking and action-taking systems" and that agent architectures are becoming a strategic transformation tool in institutions.

Three fundamental barriers in digital transformation: talent, budget, organizational culture
TRAI research shows that the areas where institutions struggle most in their artificial intelligence transformation are:

-Talent shortage (58%)

-Budget insufficiency (57%)

-Organizational culture (55%)

The support areas institutions need most are highlighted as "identifying the right use cases", "managing the implementation process" and "organizational adaptation".

On the infrastructure side, the vast majority of companies use Microsoft Azure, while Google Cloud and AWS are also among the commonly preferred platforms. The increase in hybrid structures demonstrates the adoption of multi-system usage in data and generative artificial intelligence infrastructures.

"Turkey's potential is great, but visionary leadership is needed"
TRAI Founder Halil Aksu, in assessments regarding the research findings, noted that artificial intelligence offers a historic transformation opportunity for Turkey and stated:

"Artificial intelligence is no longer just a technology; it is the language of the future, the engine of economies, the new measure of competition. Institutions in Turkey are aware of this transformation, but we have areas for development in strategy, competence and culture. While this report holds up a mirror to Turkey's artificial intelligence journey, it is also a call to action. Our potential is great; but to turn this potential into reality, we need visionary leadership, strong partnerships and decisive steps."