Back in October 2023, during the keynote address at the leading annual financial services conference in Luxembourg, going against the AI hype, I urged attendees to remain cautious about AI investments, advising them to experiment before they commit—at least not without considering the long-term implications for safety, governance and cost.
I anticipated that business models would change and costs would eventually decline, and warned that the promise of cutting-edge AI at a premium might soon feel like a poor bet. Eighteen months later, the landscape has shifted more rapidly than anyone could have predicted and confirmed my prediction.
Indeed, a cautious AI strategy has proven prescient, and those who heeded my advice are now reaping the rewards. I know this to be the case, as they have shared their gratitude. I take satisfaction in knowing that my guidance helped preserve their careers, reputations, and substantial AI budgets, which remain firmly intact in their accounts ready to be spend on what moves the needle .
While 2023 and 2024 were years of restraint—marked by experimentation and preparation—2025 heralds a shift. It’s time to accelerate.
Here are several key recommendations for business leaders seeking to protect their companies amidst the rapid developments in AI technology and competition (also download the checklist at the end of the article)
Reassess AI Investments: As AI development costs continue to fall, companies must take a closer look at their existing AI investments, particularly long-term, high-cost contracts. Business leaders should consider scaling back on expensive proprietary solutions and explore more cost-effective, open-source models, such as those championed by firms like DeepSeek.
Remain Agile in Technology Adoption: The AI landscape is shifting swiftly. Leaders should foster a culture of experimentation, as advocated in the article, to stay ahead of disruptive technologies. This entails avoiding premature, long-term commitments and instead piloting AI solutions as they emerge. Such flexibility is key to responding to new entrants like DeepSeek, which could swiftly alter the competitive dynamics.
Prioritise Data Security and Governance: With the rise of open-source AI models, companies must reassess their data security frameworks. The growing use of DeepSeek’s AI models, particularly outside the US, brings important questions of privacy, security, and compliance. Business leaders should establish comprehensive governance structures to safeguard sensitive data, especially as reliance on foreign-developed AI systems increases.
Diversify AI Supply Chains: As AI development becomes more decentralised and led by companies with fewer resources, leaders should consider diversifying their AI supply chains to mitigate risks associated with over-reliance on specific hardware or software providers. DeepSeek’s rise shows that AI development is no longer solely dominated by industry giants like NVIDIA or OpenAI. Exploring alternative AI models and hardware can provide a competitive edge and mitigate risks tied to global supply chain disruptions.
Prepare for Intensified Competition and Margin Pressure: The trend towards more affordable AI systems signals heightened competition, which could squeeze margins for firms that depend on high-cost, resource-heavy models. Leaders should prepare for this shift by seeking efficiencies within their own operations, exploring ways to leverage AI for enhanced productivity without inflating costs.
Monitor Geopolitical Dynamics: The geopolitical environment will play a significant role in the future of AI. The emergence of Chinese AI companies such as DeepSeek, and the US’s efforts to control the export of high-end semiconductors, underscore the growing influence of national policies on the tech sector. Business leaders should stay vigilant to the regulatory landscape and be prepared for shifts in the geopolitical balance that could affect access to AI tools and data.
Reevaluate Hardware and Infrastructure Needs: Innovations in hardware, particularly NVIDIA’s new personal computing devices, may make powerful AI development more accessible. Leaders should consider reducing reliance on traditional data centres and exploring more cost-effective ways of deploying AI, especially as smaller players and open-source models gain traction.
Review Long-Term AI Strategy: Leaders must carefully evaluate their long-term AI strategies, particularly given the changing cost trajectory. Ensuring that AI investments align with the company’s broader strategic objectives is essential to avoid overcommitting to outdated or overpriced technologies. This includes understanding how AI can be effectively integrated into existing business models to drive growth and improve efficiency, while avoiding unnecessary expenditure. The leanest developments are pointing to an expectation of continued rapid acceleration of AI capabilities. Companies need to articulate clearly what that looks like to them, what does it mean to their business models and how can they innovate new business models, revenue models and profitably models.
Comments