There is a growing connection between generative AI (GenAI) and hybrid cloud adoption and integration. In an article, IBM posits that GenAI can build a successful hybrid cloud strategy by meeting complex market demands. Moreover, hybrid cloud and adopting GenAI go hand-in-hand to promote sustainability objectives of enterprises.
In a recent interview with FutureCIO, Rodney Regalado, the country manager of Infrastructure Technology at IBM Singapore, discusses challenges in the deployment of GenAI in hybrid cloud environments and how to overcome them, its impact on industry-wide practices, and bridging the skills gap in terms of its deployment.
Not without a challenge
Regalado observes that organisations have accelerated the adoption of GenAI in a hybrid cloud environment, with around 66% of hybrid cloud adopters in Singapore already establishing policies to deploy GenAI.
However, some challenges and requirements remain, including the ability to control and manage the hybrid cloud environment, the need for a flexible and portable platform, and an increased agility with the ability to connect cloud and on-premises environments.
“When it comes to deploying generative AI within a hybrid cloud environment, enterprises are also concerned about the potential exposure of sensitive data,” Regalado says.
Around 41% of cloud leaders in Singapore are still concerned about cybersecurity and privacy, based on an IBM study.
Overcoming cybersecurity and privacy challenges
“Generative AI can fortify business defence – speeding up security processes that were once a heavy lift and quickly analysing vast amounts of data to spot anomalies and threats as they materialize,” says Regalado.
However, malicious players also use the same technology that fortifies cybersecurity to upgrade and sophisticate their attacks.
“With large language models (LLMs) designed to generate realistic outputs, it can be challenging for unsuspecting users to discern incorrect or malicious information. This dynamic is amplified in a hybrid cloud setup with its multiple entry points in both on-premise and public cloud environments,” Regalado says.
He adds that organisations should prioritise data policies and controls centred on security, privacy, governance, and compliance.
“Start by classifying sensitive data for enhanced protection and prevention of data loss or leakage. Subsequently, build controls around applications accessing data and enforce policies around machine learning to guard against inappropriate content,” Regalado suggests.
He shares that IBM uses IBM filters for Hate, Abuse, and Profanity (HAP) and Personal Identifiable Information (PII). The company also published details of its training data sets for its Granite models.
There’s a need to support open AI innovation. AI should be built by and for the many, not the few. To that end, an open AI ecosystem is good for healthy competition, innovation, skilling, and security.
Rodney Regalado
Responsible AI use
“Organisations that want to employ generative AI to unlock new value and insights have a fundamental responsibility to foster trust in the technology. For businesses, one hallucination is one too many,” Regalado says.
He adds that governments globally are currently talking about regulating AI technology to protect users and enterprises against cyberattacks and other potential risks. The IMDA in Singapore also deployed the AI Verify Foundation, which maximises the global open-source community for responsible use of AI. Moreover, IMDA recently launched a Generative AI Evaluation Sandbox to support the initiative.
IBM also joins the mission through watsonx.governance to help organisations apply AI responsibly and prepare for regulation. To assist businesses in managing AI and preparing for broad AI regulation, the company is releasing an integrated platform with watsonx.governance.
Responsible AI cannot be an afterthought. Currently, fewer than 60% of organisations are prepared for AI regulation, and more than half are delaying investments until they have clarity on AI standards and regulations.
Rodney Regalado
Bridging the skills gap
Cloud skills remain a challenge for 54% of decision-makers, according to statistics by IBM. Regalado says organisations “should invest, continue to build and develop skills on the open-source platform as generative AI continues to evolve. The market, in general, will move in this direction given the shortage of skills across industries.”
AI integration
As integration of GenAI within a hybrid cloud environment impacts business operations and shapes the future of hybrid cloud technology, enterprises can find value in navigating its complexity of the trend to maximise its benefits.