Experts have warned that strong cybersecurity safeguards are needed to protect sensitive information as artificial intelligence plays a growing role in health care and education.
The UAE is expanding AI use across essential services as part of a wider drive to develop the technology. Abu Dhabi’s government digital strategy for 2025 to 2027 is backed by Dh13 billion ($3.54 billion), while the Ministry of Education has developed an AI literacy framework covering kindergarten to Grade 12.
In May, the UAE Cabinet also approved a national policy to advance digital health care and AI, including developing medical systems and preparing healthcare workers to use the technology.
But information such as patient records, personal data and financial details could be at risk when shared with AI tools such as ChatGPT or Claude.
Nisha Sanghani, partner and head of Middle East at Ashurst Perkins Coie Risk Advisory, said the acceleration in the use of AI must be met with significant cybersecurity investments to keep sensitive information safe, and to train people using these tools to avoid breach risks.
“As adoption accelerates, safety, governance and accountability need to advance at the same pace as innovation – particularly where AI may influence decisions about a person’s health, development or future,” said Ms Sanghani.
“In health care and education, the measure of AI success cannot simply be how quickly we adopt it, but whether it delivers better outcomes that people can trust.”
Opportunity for the UAE
Ms Sanghani said the shift from experiments to everyday services increased the importance of oversight, but that the UAE had an opportunity to demonstrate how AI could be used responsibly on a large scale.
“For organisations operating here in the UAE, the question is moving beyond ‘where can we use AI?’ to ‘where can AI genuinely create better outcomes, and what needs to be true for us to trust it?”
Nalla Karunanithy, chief executive of digital health and e-commerce at Aster DM Healthcare told The National the organisation was using AI to read X-rays alongside radiologists, measure cardiac MRI scans and supports early detection of coronary artery disease.
This helps save time for doctors so “the clinician's time goes towards judgment,” said Mr Karunanithy. “The AI detects, measures or flags; the clinician interprets, decides and signs.”
But as AI becomes more widely used, healthcare providers must control which systems can access medical information and what happens to it afterwards.
Aster stores patient information in the UAE and assesses each AI system before allowing it access, Mr Karunanithy said.
Testing is not done with real patient data and only approved systems receive necessary information.
“Our group AI governance policy prohibits external providers from using patient information to train their models or for any purpose beyond the approved scope,” he said.
UAE healthcare provider NMC Healthcare is centralising information across its 70 centres, said Ahmed Hafez, director of solution engineering for the Middle East, Turkey and Africa at data company Snowflake.
He added that the work was intended to improve understanding of hospital operations and support decisions at the point of care.
Educate the workforce
Vladislav Tushkanov, research and development group manager at cybersecurity company Kaspersky, told The National at the AI Everything summit in Abu Dhabi that personal data had to be kept safe.
“We need to understand how we educate our workforce, how we make them understand that it's OK to use the internal AI systems to process information, but you cannot send personally identifiable information in the cloud,” he said.
Aster said it had rejected solutions that could not meet its data residency requirements. Staff must use approved enterprise tools, with its AI Centre of Excellence overseeing their assessment and adoption.
Last year, hackers claimed responsibility for a large-scale data breach at American Hospital in Dubai, reportedly seizing 450 million patient records.
The hospital was contacted for comment but had not responded at the time of publication.
However, despite the possible consequences of mistakes, Costi Perricos, Deloitte’s global generative AI leader, said that this should not prevent organisations from using AI.
“You shouldn't turn away just because we need to be more careful about this,” he said.
He added that the Middle East’s access to capital and technology, combined with its willingness to pursue new ideas, raised opportunities to develop the AI sector.
Introducing tools is only one part of the work, Mr Perricos said. Organisations also need to prepare employees for changes to how they did their jobs.
Bikram Mukhi, chief strategy officer at Soika Labs, said human oversight remained critical, particularly in education.
“We live in very interesting times where technology is moving faster than the rate of adoption.”
Users need to understand that the technology is still developing and that new capabilities could bring new risks, he said, adding that governments and regulated businesses are also seeking greater control over the information and infrastructure used by their AI systems.
“AI just can't be just deployed across the board,” said Shadi Hatoum, regional manager for the Middle East and Africa at data platform company Tealium. Different uses require different standards and measures of success, he said.