Can AI fundamentally disrupt the traditional enterprise software model?
Recent share market volatility has hit software stocks as investors question whether Artificial Intelligence (AI) could fundamentally disrupt the traditional enterprise software model. Some analysts suggest that as generative AI tools become more capable of producing working applications with minimal developer input, organisations may increasingly build their own software rather than purchasing it from vendors.
However, history suggests enterprise software has never been valued for code alone. During the rise of open-source software in the 1990s, similar predictions were made about the collapse of proprietary software pricing. Instead, enterprise software vendors continued to thrive because organisations ultimately pay for reliability, governance, security, compliance and trusted outcomes.
Today, as organisations experiment with AI-generated applications, a new question is emerging: will AI truly replace enterprise platforms, or will it make the need for structured, governed software environments even more important?
The AI News Blog interviewed Medhat Galal (pictured), Senior Vice President of Engineering, Appian, to find out more.
1. Software stocks have come under pressure as investors question whether AI could undermine the enterprise software model. Do you see generative AI as a genuine threat to traditional software platforms?
Generative AI is clearly a significant shift for all enterprise organisations. There’s a perception that if AI can generate software, it will reduce coding costs and the need for enterprise platforms. In reality, that assumes the value of software sits in the code itself, which has never really been the case.
Enterprise organisations invest in software because they need reliable, secure and well-governed systems that support critical operations. Code is only one part of that equation. What matters more is how systems perform, how they are managed over time and whether they can be trusted to deliver consistent outcomes.
AI is a powerful capability, but it is inherently probabilistic. It produces outputs based on likelihood rather than certainty, which means variability is built into how it operates. That makes it highly effective for augmentation, but less suited to operating independently in environments where consistency and accuracy are essential.
So rather than displacing enterprise software, AI is increasing the importance of platforms that can provide structure, governance and control around how it is used.
2. Some analysts argue AI will allow organisations to build their own applications rather than buying software. How realistic is that for large enterprises?
AI will make it easier for organisations to build applications, particularly for narrow or low-risk use cases. But for core systems, the requirements are very different.
Large organisations need systems that integrate across functions, operate reliably at scale and meet strict governance and compliance standards. AI-generated applications do not inherently provide those capabilities.
They also lack a broader support structure. Enterprise software comes with vendor accountability, ongoing updates and access to expertise. AI-generated applications are typically bespoke, with no external support model or shared knowledge base to rely on.
There is also a practical consideration. Many organisations already operate across fragmented systems and disconnected data environments. Building additional applications in isolation can increase that complexity rather than resolve it.
This is why we are seeing organisations take a different approach in practice. For example, organisations are focusing on improving how processes operate across systems and teams, rather than simply introducing new tools in isolation. By strengthening process foundations, they are able to apply AI in a way that delivers more consistent and scalable outcomes.
This reflects a broader shift towards applying AI within end-to-end processes rather than as standalone tools, ensuring outputs are connected, governed and actionable.
3. History offers an interesting parallel in the rise of open-source software in the 1990s, when many predicted proprietary software would lose its value. What lessons from that era are relevant to today’s AI debate?
One of the clearest lessons is that making code more accessible does not reduce the need for enterprise-grade software.
Open-source software significantly expanded access to code, and at the time there were similar concerns that freely available code would undermine the software industry. However, it did not diminish the importance of reliability, support and governance. In many ways, the opposite happened and it shifted attention toward those factors.
AI is creating a similar dynamic. It is accelerating how software can be developed, but it does not address the operational requirements that organisations depend on, such as stability, compliance and long-term maintainability.
There is also an important distinction. Open-source software was supported by broad communities contributing knowledge, updates and problem-solving. AI-generated applications, by contrast, are often built for a single organisation, with no external support network or shared expertise to rely on if issues arise.
4. Enterprise platforms have traditionally provided governance, security and reliability around critical processes. How difficult would it be for organisations to replicate that with AI-generated applications?
Replicating that level of governance is extremely challenging without a structured platform.
Enterprise systems are designed to manage how work flows across an organisation. They provide visibility, enforce controls and ensure that processes are executed consistently. These capabilities are particularly important in environments where compliance and operational risk must be carefully managed.
In Australia, regulatory expectations are increasing. Frameworks such as APRA CPS 230 (covering banking, insurance and superannuation) are placing greater emphasis on operational resilience, service continuity and the ability to trace how decisions are made.
These requirements highlight the need for systems that can provide transparency and accountability. AI-generated applications, on their own, do not offer that level of oversight.
Embedding AI into end-to-end processes allows organisations to benefit from its capabilities while maintaining the governance and control required in enterprise environments. This is particularly evident in sectors such as banking, where institutions are prioritising process visibility and control to strengthen compliance outcomes while scaling new technologies.
5. There is growing excitement around “AI building software”. But in highly regulated industries such as banking, insurance or healthcare, how feasible is it to rely on AI-generated applications without strong oversight and governance?
In regulated industries, oversight and governance are not optional. Systems must operate in a way that is transparent, auditable and aligned with regulatory expectations. Many enterprise processes require a high degree of accuracy and consistency. Even small errors can have significant consequences, which makes it difficult to rely on AI in isolation.
That’s why many of Appian’s customers across the globe sit within high regulated industries such as government, financial services and healthcare.
One example is Acclaim Autism, which treats children with Autism Spectrum Disorder. Their manual intake process left patients waiting up to six months for care. With Appian, they cut intake time by 83%. Acclaim Autism implemented Appian AI in just three weeks to extract diagnosis information from unstructured medical documents, ensuring accurate medical information is reflected to guarantee compliance and expedite the intake process. And because of Appian’s private AI approach, they remain compliant with HIPAA and other privacy standards. Now, they are delivering faster care to patients in need.
6. Where do you believe AI will deliver the most value within enterprise software platforms: accelerating development, automating processes, improving decision-making, or something else?
AI delivers the most value when it is applied within processes rather than as a standalone capability.
One of the most significant opportunities lies in improving visibility into how work is actually performed. Process intelligence allows organisations to analyse workflows, identify inefficiencies and understand where delays or rework occur. By combining that visibility with AI, organisations can make more informed decisions, automate routine activities and continuously improve how work is executed.
This approach is already delivering results across industries. In financial services, Australia’s Westpac Banking Corporation is utilising generative AI across different areas of the business to make it easier for their people to deliver better and faster results for customers. Westpac’s Mortgage Assessor AI solution uses the Appian platform to help mortgage assessors to support brokers, reducing the time to assess loans and speeding up approval times.
7. Looking ahead five years, do you expect AI to reduce the role of enterprise platforms, or will it actually reinforce the importance of trusted systems to manage complex business operations?
AI is likely to reinforce the importance of enterprise platforms rather than reduce it. As organisations rely more heavily on AI, the need for systems that can ensure consistency, governance and accountability will become more pronounced.
Trust will be a critical factor. Organisations will need to ensure that AI operates within defined parameters, produces outcomes that can be explained and aligns with regulatory and operational requirements. That requires platforms that can connect data, manage processes and provide clear oversight of how decisions are made.
In that sense, AI is not replacing enterprise software. It is raising expectations for what enterprise systems need to deliver. The organisations that succeed will be those that combine AI with strong process foundations and governance, enabling them to scale innovation without compromising control.
