Organisations in the Middle East and Africa are taking a more disciplined approach to adopting artificial intelligence to prioritise control of data and "operational control", a study has shown.
The research by Veeam Software found that 60 per cent of businesses and organisations surveyed in the Middle East and Africa region considered data sovereignty to be a top factor in deciding how to adopt AI in their workflows.
That proportion is higher than the global average of 56.6 per cent, Veeam Software's analysis showed.
The survey included responses from 1,000 people working in enterprise IT, data and security around the world.
Although definitions of data sovereignty can vary, in the context of AI and technology the term refers to entities maintaining control over proprietary information and data.
"Organisations across the Middle East and Africa increasingly recognise that data sovereignty is not simply a compliance exercise," said Mena Migally, a regional vice president at Veeam.
"It is a strategic enabler for building trust in AI and driving secure digital transformation."
The analysis also found that, in terms of carrying out plans to maintain data sovereignty in light of AI adoption, the Middle East and Africa region was the "most mature".
"The challenge now lies in maintaining visibility and control across increasingly complex ecosystems that span cloud environments, AI platforms and third-party providers," Mr Migally said.
The continuing AI boom has prompted debate over the sanctity of data, and, more specifically, what should and should not be uploaded to AI chatbots.
During an interview at the World Economic Forum's Annual Meeting in Davos, Switzerland, in January, Sandbox AQ chief executive Jack Hidary reflected on people uploading sensitive information to various AI tools without thinking of the consequences.
“I'm not talking about closed, licensed AI tools, I'm talking about generic chatbots like ChatGPT and Gemini,” Mr Hidary told The National.
He pointed to a recent incident involving Samsung when two engineers put a prompt into ChatGPT using proprietary information, and this "became part of the data that you can query today".
"Similar incidents are happening every day," he said.
Veeam's survey results arrive as AI spending comes under closer scrutiny, with executives looking for a return on investment.
That scrutiny is also causing debate among the founders and executives of AI companies jostling for influence and seeking the upper hand in promoting their business models.
In a recent interview on CNBC, Palantir chief executive Alex Karp said his company was in the perfect position to capitalise on concerns about data sovereignty, compared to competing AI firms such as OpenAI and Anthropic.
"I'm not throwing shade at them but something has gone completely wrong," he said, alleging that many in the private sector were tired of spending vast amounts of money on AI tools that don't necessarily protect data.
Palantir's Ontology system, he said, is safe "because it doesn't touch your underlying data, safe because it prevents the large language model from capturing your data and replicating your business, and is safe because it doesn't transfer intellectual property".
