2025 Predictions: Generative AI continues to set the stage for a transformative year
By Keir Garrett (pictured), Regional Vice President, Cloudera, Australia and New Zealand
As we look forward to 2025, the excitement around Artificial Intelligence (AI) and Generative AI (GenAI) is palpable, as it drives innovation across organisations. However, this initial headrush is likely to wane as businesses shift their focus to the practicalities of investing in AI and measuring its effectiveness. The Australian landscape is evolving rapidly as well, with legislative reforms and new regulatory policies shaping the future of governance, cybersecurity and ethical AI use, significantly impacting business operations.
For organisations, effective data management is no longer optional – it is essential for compliance and for showcasing the tangible benefits of AI investments.
Cloudera identifies five key shifts that businesses in ANZ should be aware of to navigate these changes and unlock long-term value from AI technologies in the year ahead.
Businesses seek measurable results from AI investments
By 2025, many businesses will have successfully harnessed the benefits of GenAI, with 65% of organisations already reporting regular use to achieve significant cost savings in HR and revenue increases in supply chain management, as reported by McKinsey. Despite this, the initial excitement is wearing off as technology leaders grapple with the reality of scaling AI from pilot projects to fully integrated solutions with measurable outcomes. In fact, at a recent summit of ANZ IT leaders, 70% of attendees reported being at various stages of AI implementation yet also acknowledged a skills and confidence gap.
Bridging the AI confidence gap remains a formidable challenge as many organisations are still grappling with their foundational data management and achieving a ‘single source of truth’. This becomes a bigger problem for those organisations without large data pools which can lead to fragmented insights and inefficiencies.
To succeed in AI, organisations must adopt a holistic view of data, treating it as a core asset that drives decision-making and innovation. This involves breaking down data silos and fostering a culture of data collaboration across departments. Advanced data platforms play a pivotal role here, enabling seamless integration of disparate data sources and ensuring consistency and accuracy.
A hybrid cloud infrastructure is no longer sufficient
If 2024 was the pilot year for Gen AI, 2025 will focus on scaling. Gartner predicts that cloud will remain the top platform for GenAI-enabled applications, with Australian IT spending expected to hit AU$147 billion, driven in part by investments in cloud, including AU$23 billion on public cloud services in 2024 alone.
Many organisations in Australia are still in cloud adoption mode – often operating under the mindset of, “Why build from scratch if we can get it off the shelf?” Additionally, there is a growing trend of developing niche AI solutions tailored to specific workloads, such as automated customer service response systems to optimise efficiency and reduce costs.
As we move in 2025, organisations must embrace a comprehensive strategy to cloud adoption that prioritises flexibility, security and innovation. This approach will enable them to effectively harness the power of their data while navigating the complexities of a multi-cloud environment.
To achieve this, businesses will need robust multi-cloud capabilities to manage data across on-premises, mainframes, public cloud, and edge environments.
AI Assistants will play a crucial role in accelerating cloud adoption by streamlining data management and enhancing collaboration. Hybrid data management platforms will become essential for integrating diverse data sources while maintaining control and security over operations.
Businesses will favour private LLMs over public LLMs
Enterprise AI innovation is driving a shift towards enterprise-grade private LLMs for tailored insights. According to a McKinsey study, less than half (47%) of companies are significantly customising and developing their own models currently and we believe that this is set to change in 2025 as businesses develop AI-driven chatbots, virtual assistants, and agentic applications tailored to the individual business and industry.
In Australia, there is currently a preference towards public LLMs due to the convenience and cost factors. However, as the stakes rise with significant reputational, financial, legal and customer retention risks, highly regulated industries will lean towards private LLMs. This shift is driven by the need for enhanced security and customisation, similar to how the government is actively promoting operational resilience through standards like CPS 230.
As enterprise-grade LLMs become more common, they will require superior GPU performance and robust data governance systems for improved security and privacy. The adoption of retrieval-augmented generation will also grow, transforming generic LLMs into industry-specific data repositories, enhancing accuracy and reliability for end users in areas such as field support, HR, and supply chain management. To truly harness the power of AI, businesses must invest in building and customising LLMs to align with their specific needs to ensure precise and impactful outcomes.
The rise of FinOps as a discipline to optimise costs
The accelerated adoption of AI will drive an unprecedented demand for compute power, essential for training and running machine learning (ML) models. This surge will lead to a surge in operating expenses with Gartner suggesting that improving operational margins will be the most critical outcome from technology investments for 94% of ANZ CIOs next year, up from 48% in 2024.
We see FinOps becoming hugely important for technology leaders to balance AI innovation with the need to achieve greater cost efficiencies. Additionally, cost optimisation should prioritise achieving faster time to value and enabling better business decisions. AI and ML offer powerful solutions to manage these pressures. AI-powered analytics enable real-time monitoring and management of cloud costs, identifying cost-saving opportunities and optimising resource allocation. Similarly, predictive models will help forecast future cloud usage and costs, allowing organisations to anticipate demand spikes and plan resources accordingly.
The integration of FinOps and AI introduces a new dimension to innovation and cost optimisation. Organisations will require modern, cloud-native platforms to seamlessly transition data and workloads across different cloud environments, maximising the return on cloud investments and positioning businesses for sustainable growth in an increasingly competitive landscape.
The regulatory landscape will continue to evolve to reduce risk and improve trust
With Australian legislative reforms on compliance, cybersecurity, data privacy and ethical AI, significant changes in data security and governance are expected. Gartner forecasts that by 2028, organisations with comprehensive AI governance platforms could see 40% fewer AI-related ethical incidents.
While compliance costs are high, these investments are crucial for sustainable growth and trust. Business leaders must champion data governance frameworks that uphold data quality and integrity while investing in advanced data management solutions to enhance security, ensure data integrity and navigate the regulatory landscape effectively.
