Why AI is strategic but still immature in businesses

Let's start from the end, from what is a well-known assumption and one that finds increasing consistency in the now almost redundant discussion around artificial intelligence: AI is becoming more and more central to corporate strategies, but the path marking its adoption is far from simple amid technical, ethical and operational difficulties. Algorithms, generative tools, LLM models and agents, this is the message that must pass and take root, are not just advanced technologies, but constitute (if understood and well deployed) a strategic lever that accelerates innovation, reduces costs and creates competitive advantages. The photograph taken by the AI Pulse Survey, the new global observatory of Protiviti, a multinational management consulting group, confirms in this sense how the corporate ecosystem is experiencing a phase of strong dynamism on the artificial intelligence adoption front, but with still fragmented levels of maturity. The study is composed of three different reports ('From Exploration to Transformation', 'From Data Confusion to AI Confidence' and 'From Automation to Autonomy') and reflects the perceptions of more than a thousand professionals and managers active in all the main business functions and in a wide representation of industries, offering a detailed reading of how artificial intelligence is transforming operational models, decision-making structures and data governance in companies.

AI maturity: less than one in ten companies are at an advanced stage

The first report analyses the level of maturity, perceived benefits and future prospects of AI in the corporate environment and highlights how the ongoing transformation path is far from uniform. Only 8% of organisations, in fact, reach an advanced stage of execution, where artificial intelligence is strategically integrated and generates competitive advantages, while the rest of the corporate world is distributed between exploration, experimentation and first operational applications. Yet the return on investment is clear: 85 per cent of organisations report a positive ROI, with over 47 per cent of the most mature organisations (partly due to a more robust and structured data infrastructure) claiming to have 'significantly exceeded' expectations. The AI adoption path, according to Protiviti's experts, unfolds through five different levels of maturity: from initial exploration to advanced transformation, passing through intermediate stages of process definition, prototyping and integration. And each of these levels is characterised by a different relationship between people, processes and technology. It starts from the discovery and experimentation phase through small use cases and without a defined strategic framework to a second step where technology enters operational functions, but remains confined to uncoordinated initiatives. At the third stage, the first integrations of the technology into workflows begin, with tangible benefits especially in terms of productivity. In the fourth stage, AI then becomes an integral part of decision-making processes to assume (in the fifth stage) the status of a strategic asset embedded in the company's operational core.

As things stand, most companies are concentrated in the first three tiers (where the prevailing benefits relate to savings and efficiency) and only a minority, less than one in ten, achieve full integration between the potential of technology, data governance and the ability to scale projects, generating benefits ranging from increased revenues to reduced project time-to-value and increased competitiveness. The gap between those who are ahead and those who are currently lagging behind, the report goes on to say, is not only technological: more mature companies demonstrate a more robust data culture, more structured evaluation and monitoring processes and a superior ability to measure the impact of AI on key business indicators.

Data quality as an enabling (and discriminating) factor