The rise of artificial intelligence in software development is reshaping the commerce landscape, offering faster product launches and more personalized customer experiences.
Qodo’s $40 million Series A funding round in September for its AI-powered code testing platform reflects growing investor confidence in AI’s ability to streamline business operations and enhance digital services.
Coding tools like GitHub Copilot and OpenAI’s Codex are changing how companies build and deploy software. These advanced machine-learning models can suggest code snippets, complete functions, or create entire code files based on prompts or existing code.
“AI coding tools enhance the productivity of developers greatly through the automation of some repetitive tasks and code suggestions,” Dhaval Gajjar, chief technology officer of Textdrip, a Software-as-a-Service (SaaS) company, told PYMNTS. “This can bring about faster development cycles and, consequently, reduce the time-to-market.”
These tools “maintain the quality of code based on best practices and catch potential errors right at the development stage,” Gajjar said. “It reduces an extended testing and debugging process, thereby saving a lot of time and resources.”
The impact goes beyond productivity gains. Amazon CEO Andy Jassy highlighted the effect of the company’s AI-powered code transformation capability, Amazon Q, on social platform X.
“The average time to upgrade an application to Java 17 plummeted from what’s typically 50 developer-days to just a few hours,” he posted. “We estimate this has saved us the equivalent of 4,500 developer-years of work…”
These efficiency gains could reduce development costs and timelines across various industries, accelerating innovation and time-to-market for new features and products.
The power of AI in software development is particularly pronounced in eCommerce.
“In the eCommerce space, tools like GitHub Copilot and Cursor are proving particularly valuable for rapidly implementing standard features,” Dev Nag, CEO of QueryPal, an eCommerce solutions provider, told PYMNTS. “They excel at generating boilerplate code for product catalog structures, basic shopping cart functionality and user authentication flows.”
AI-generated code offers advantages in personalization and customer experience.
“AI-generated code can easily go through large datasets containing customer preferences and behavior quickly,” Gajjar said. “For example, one can easily generate a product recommendation using AI by just tracking a user’s past purchases and browsing history.”
The technology also promises improved transaction security.
“AI can also be used to generate adaptive security algorithms that detect and prevent fraud in real time,” Gajjar said. “For example, an AI tool would give a code for a payment gateway so that a fraction of the transactions will automatically raise red flags based on the established fraud patterns, just like how PayPal or Stripe use AI for fraud detection.”
Integration of AI in software development brings challenges.
“There have been cases where AI-generated code introduced subtle bugs in inventory management systems, leading to overselling or stockouts,” Nag said.
Denisse Damian, an AI researcher, sounded another alarm.
“The rise of hyper-personalization threats is a concern,” she told PYMNTS. “Scammers could use AI to generate realistic customer service voices or emails, tricking customers into divulging sensitive information or making fraudulent purchases. With AI-generated code creating tailored digital experiences, the line between legitimate personalization and malicious exploitation could blur.”
These risks underscore the need for human oversight.
“The biggest risk businesses face with AI-powered coding tools is when engineers rely too heavily on them without thoroughly reviewing the output,” Damian said. “AI can sometimes generate code that looks correct but contains bugs or security flaws. If developers don’t catch these issues and trust the AI blindly, they could introduce serious vulnerabilities into proprietary systems.”
Gajjar outlined risks related to proprietary technology and cybersecurity.
“AI models trained on proprietary codebases could end up learning sensitive information that the model replicates, thereby exposing the system to unauthorized access,” he said, adding that there are risks associated with depending on third-party AI technology and potential supply chain vulnerabilities.
The industry may see further specialization in AI tools.
“We’re likely to see more eCommerce-specific AI coding assistants,” Nag said. “These could be trained on specialized eCommerce frameworks and best practices, making them even more valuable for the industry.”
He also had a warning, however.
“This specialization might also increase the risk of homogenization in eCommerce platforms, potentially making unique, innovative implementations more valuable than ever,” he said.
For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.
We’re always on the lookout for opportunities to partner with innovators and disruptors.
Learn More