GenAI not production-ready?

Databricks has unveiled a new Economist Impact report, “Unlocking Enterprise AI: Opportunities and Strategies,” which examines the challenges businesses face in adopting and scaling AI, and the techniques they are using to drive greater value from these investments. The report found the vast majority of global enterprises (85%) are using or testing generative AI (GenAI) in at least one function - and this rises to 90% of UK enterprises. However, just 22% of global enterprises feel confident that their current IT architecture could effectively support new AI applications moving forward. Moreover, 60% of UK enterprises admit that GenAI use cases have not yet made it into production internally.

As demand for data intelligence grows worldwide, AI continues to be a major focus area for companies. According to Goldman Sachs, global AI spend is expected to reach $1 trillion in the next few years. While more companies are investing in AI than ever before, struggles related to delivering business-specific, highly accurate, and well-governed results at a reasonable cost are preventing organisations from scaling their AI efforts and achieving more transformational results. Today, only 37% of executives globally believe their GenAI applications are production-ready. This figure falls to just 29% among practitioners, who cite key hurdles including cost (41%), skills (40%), quality (37%) and governance (33%).

“Whilst it’s encouraging to see so many UK enterprises already using or testing GenAI, the fact that so few feel their architectures are ready for this technology is notable. There is undoubtedly enthusiasm for everything that GenAI can help enterprises achieve, but still a myriad of barriers to overcome first,” said Michael Green, VP of Northern Europe, Databricks. “It therefore comes as little surprise that enterprises are seeking solutions that prioritise data, centralise governance and deliver TCO at scale - all whilst being tailored to the organisation. This report from Economist Impact showcases why data intelligence is essential, particularly as a solution to GenAI adoption challenges - enabling organisations to take a holistic approach to data management and governance, creating the perfect environment for GenAI use cases to flourish.”

Whether streamlining clinical trials or identifying potential vehicle issues before they occur, many enterprises are already using AI to improve efficiency and productivity. With the growth of ‘Agentic AI’ — artificial agents with a natural language interface that can plan and execute tasks on behalf of a user — companies can spread these benefits to more of the workforce. In fact, nearly 60% of respondents expect that, within the next three years, natural language will be the primary or only way non-technical staff will interact with complex datasets. Increasingly, organisations are also using AI to improve customer service, fraud detection and patient care, among the many other use cases, highlighting the long-term potential of the technology to accelerate overall business success.

“AI can lead to gains in productivity across the workforce. And for businesses just starting out on their AI journeys, it’s a logical way to measure initial progress,” said Senthil Ramani, Global Lead, Data and AI at Accenture. “However, organisations aiming to become the AI leaders of tomorrow will need to capitalise on the use of the technology to drive growth, enhance customer experience, manage risk and unleash enterprise knowledge. This holistic approach will not only boost efficiency but also open new business opportunities and can attract and retain talent.”

The Economist Impact report surveyed 1,100 technical executives and technologists from 19 countries across Asia, Europe and the Americas and included additional insights from 28 C-Suite executives from 11 industries. Among the organisations represented are Accenture, CJ CheilJedang, Condé Nast, Dream Sports, Fanatics Betting & Gaming, Flo Health, Frontier, General Motors, HP, JetBlue, Mahindra Group, Mastercard, Molson Coors, Novartis, NTT Docomo, Opendoor, Providence, Rakuten Group, Repsol, Rivian, Seven West Media, Shell, Siam Commercial Bank, TD Bank Group, Thermo Fisher Scientific, Unilever, UPS and the United States Army.

Additional key findings include:

• Only 11% of UK respondents believe AI is overhyped. In fact, 83% see the technology as crucial to their long-term goals. Despite the momentum, only 27% believe investment across technical and non-technical domains is sufficient.

• Large organisations are flocking to GenAI, with 97% of companies with over $10 billion in revenue globally now using the technology in at least one internal business function. By 2027, 99% of all respondents worldwide expect GenAI adoption across both internal and external use cases.

• Globally, nearly half of data scientists (45%) are still using a general-purpose large language model (LLM) without contextual enterprise data. Those models often struggle to provide the necessary quality, governance and the ability to evaluate outputs. 58% of all data scientists have begun to augment their LLMs with proprietary data through retrieval augmented generation (RAG). In the UK, 71% of organisations see significant potential in integrating GenAI models with their own proprietary data. In fact, by 2027, 96% of UK enterprises plan to develop custom models based on proprietary data.

• Organisations expect to mix and match different models and tools in their Agent Systems, spanning open source and proprietary technologies, to drive better performance. By 2027, 96% of organisations globally plan to deploy open source AI models.

• Only half of UK respondents are confident their organisation can secure enough AI talent.

• 33% of UK respondents acknowledge their organisation’s data and AI governance is insufficient.

“From classic machine learning to generative AI, the business world’s obsession with AI isn’t letting up. But our findings show that, for many organisations, the real value comes when the technology is unleashed on their own proprietary data to develop data intelligence,” said Tamzin Booth, Editorial Director of Economist Impact. “That data intelligence is even more valuable in an increasingly unpredictable world. To drive the algorithm advantage they’re seeking, it’s clear enterprises must address significant challenges with producing high-quality outputs, identify ways to evaluate performance and governance with large AI models, and work out how to effectively connect AI to the workforce.”