The Make, Buy, Ally Decision – The Future Of Organisational AI

Hello and welcome to the latest edition of AI Strategy Brief. In this edition we’ll review some recent industry developments in AI, and consider the Make, Buy, Ally Decision leaders will face in managing their organisation’s AI investments. 

AI Industry Developments:

• Policy – New AI Trade Association Emerges
• Insurance – AXA Deploys In-House AI
• Media – Meta Launching New AI Chatbots

Policy – New AI Trade Association Emerges

In an industry defining move, OpenAI, Microsoft, Anthropic, and Google have come together to form the Frontier Model Forum, a new industry association designed to represent organisations at the forefront of AI development. The Forum states its initial purpose as aiming to ensure the safe and responsible evolution of frontier AI models, with a focus on LLMs in particular. The founding members of the Frontier Model Forum have stated their collective objectives as follows:

1. Advancing AI safety research to promote responsible development of frontier models, minimize risks, and enable independent, standardized evaluations of capabilities and safety.
2. Identifying best practices for the responsible development and deployment of frontier models, helping the public understand the nature, capabilities, limitations, and impact of the technology.
3. Collaborating with policymakers, academics, civil society, and companies to share knowledge about trust and safety risks.
4. Supporting efforts to develop applications that can help meet society's greatest challenges,

The Frontier Model Forum is a hugely significant development in the global AI ecosystem, driven by the shared ambition of its members to shape the future of AI policy and technology. I expect the output of the Forum will carry serious weight in policy conversations worldwide and is certainly one to watch. 

Insurance – AXA Deploys In-House AI

Recognising the potential security vulnerabilities that come with open-source GPT tools, AXA Group took the step to develop their own in-house GPT solution in conjunction with Microsoft. The in-house GPT platform, named AXA Secure GPT, reflects the organisation's desire to balance technological innovation while also prioritising safety. AXA's approach focuses on their platform's security and integrity, with generated content carefully managed to meet the company's governance and operational needs. 

There are several benefits to AXA's In-House AI solution, contributing to its unique value proposition. Firstly, it enables greater control and compliance. The platform enables AXA to exercise comprehensive control over the content generated, ensuring accuracy, compliance, and alignment with company values. Also, the in-house format offers greatly enhanced security. With dedicated security measures, AXA's In-House AI is a more robust, secure alternative to open-source options. Leveraging Microsoft's Azure OpenAI Service, the platform is initially available to 1,000 staff within the global AXA Group, with future rollouts planned to the wider staff base.

Media – Meta Launching New AI Chatbots

In an ambitious move that signals the next step in their AI strategy, Meta is set to launch an array of AI-powered chatbots across its Facebook and Instagram platforms according to the Financial Times. Scheduled for release as early as next month, the chatbots are designed to engage users in human-like discussions, offering recommendations, search results, and customer service. This change draws on Meta's open-source AI language model, Llama 2. The upcoming chatbots will not be confined to a niche audience; and are set to be available to all Facebook and Instagram users worldwide once the roll-out is complete. Though the full details have yet to be released, Meta has unveiled some core capabilities that these chatbots will possess:

• Fielding questions about Facebook and Instagram features with ease and expertise.
• Offering tailored recommendations for content and products, providing personalised experiences.
• Acting as customer service agents, addressing issues with responsiveness and precision.
• Engaging in natural conversations with users, blurring the lines between machine and human interaction.

Accessible through the Messenger App or directly on the Facebook and Instagram websites the chatbots initiative aims to improve the user experience on Facebook and Instagram and create greater engagement on the platforms. The potential marketing ROI in this sort of engagement should not be underestimated and will represent the next stage in the evolution of business-to-consumer social media marketing. 

The Big Idea: The Make, Buy, Ally Decision – The Future Of Organisational AI

With news of AXA Group’s deployment of In-House AI, we are starting to see the beginning of the first wave of large-scale AI deployments in organisations. Over the next year we will likely see many such Corporate – GPTs rolled out, especially in industries that rely heavily on knowledge work. With the pending release of Microsoft’s productivity AI, Copilot, this represents two of the major changes we will see from AI in the short-term. But what is an In-house AI and why would a company want to develop one?

What is In-House AI?

To start with, In-house AI refers to the development and implementation of artificial intelligence solutions within a single organisation, leveraging internal resources and expertise rather than outside services. With this definition, AXA’s new AXA Secure GPT is not a true In-House AI, rather one that has been developed in conjunction with an AI technology partner, in this case Microsoft. The value proposition of In-house AI stems from the greater control it offers over the technology and how strongly it can be aligned to an organisation’s specific needs, goals, values and governance guardrails. However, the development of AI applications also requires substantial investment in talent, infrastructure, not to mention the burden of ongoing monitoring and management. This makes AI development a far from simple decision.

To Make, Buy, or Ally

Like any technology or capability in limited supply, organisations and leaders are faced with a difficult decision of how exactly to gain and develop new AI capabilities. This is the Make, Buy or Ally decision of Artificial Intelligence:

• The 'Make' Strategy: Building AI internally allows for maximum customisation and alignment with specific business objectives but requires substantial investment in expertise and infrastructure.
• The 'Buy' Strategy: Purchasing AI solutions offers speed and efficiency, enabling immediate integration, but may lack the tailored fit of an in-house solution.
• The 'Ally' Strategy: Partnering with AI providers allows for a balanced approach, leveraging external expertise while maintaining some level of customisation and control. Collaborative partnerships may also foster innovation and growth but can lock organisations in and reduce strategic freedom. 

A clear understanding of these options, and how each aligns with the business's strategic direction, is needed to guide an informed and successful AI integration path. The Make, Buy or Ally decision is something that should be at the heart of an organisation’s AI Strategy and directly impacts both the AI Alignment and AI Governance Challenge of the organisation. Given the pace of change in AI capabilities across industries, getting this decision right is only becoming harder, as new AI capabilities and technologies are quickly coming online, evolving, and improving.

The Make, Buy, Ally Decision & AI Strategy

The Make, Buy, or Ally Decision is not a simple choice by any means. To begin with, organisations will have limited resources which they can draw upon to understand the potential use cases of the technology and aid in vendor selection. Certain options such as organic development may not even be open to them. It can also be challenging to understand how the Make, Buy or Ally decision on AI will relate to the organisation's overarching strategy. If there is poor alignment between overarching strategy and AI Strategy driven investments, an organisation will likely suffer from wasted time and resources at best and deeper strategic and competitive problems over the medium-term. While this decision is a step that must be built into the strategy process of the organisation for the senior team to cycle through repeatedly, the deeper and more significant challenge of the Make, Buy, or Ally decision relates to AI Governance.

The Make, Buy, Ally Decisions & AI Governance

When it comes to sourcing AI Applications and capabilities, leaders face a great challenge in the form of the AI Governance burden it places on an organisation. Leaders should be aware of the full extent of the AI Portfolio being deployed across the organisation and the multitude of AI Governance touch points and challenges, none of which are static. Issues such as the Layering Problem have to be dealt with and continually managed before the benefits of Centaur Usage can be accrued for knowledge workers. Where AI is to be deployed in a fully autonomous role, significant governance checks have to be put in place to ensure consistency, fidelity, and technical integrity throughout operations. An organisation’s AI Governance System must deal with all of these challenges as they evolve and move over time, keeping a vast number of stakeholder considerations and positions in mind. 

The Coming SME Challenge

SMEs face unique challenges and opportunities in AI integration as well, and I do not want to leave them out of this conversation. In many ways, the challenges of adapting and integrating AI Applications and technologies are even greater for SME leaders, and yet this is precisely where AI developments will have the greatest impact on everyday life for ordinary citizens and consumers.  Tailored solutions must consider budget constraints, accessibility to specialised talent, and alignment with specific business needs before SME leaders can really commit to any AI expenditure. The costs of even frontline AI technologies, such as Microsoft’s forthcoming Copilot package, may prove too much for many SMEs, and careful decisions have to be made around who within an SME should be AI-resourced and for what purpose. 

Supportive AI Policy For SMEs

There are many positives to be gained for SMEs though, and the biggest productivity gains may be seen in this sector of the economy. Accessing AI solutions through standard ‘Buy’ offerings like OpenAI and Microsoft’s Office 365 will provide huge productivity and time saving gains to many overworked and under-resourced entrepreneurs and SME leaders. AI is often mentioned as a job destroyer, but if it can help SME leaders grow their businesses faster, AI will serve as one of the greatest engines of job creation in the modern era. Forward looking policy makers may even consider the creation of targeted AI Grant Schemes to provide SME leaders with education and support in deploying AI tools, with the view to growing their companies.

Industry-Specific Considerations

Different industries pose unique demands and regulatory environments that must be factored into AI integration strategies. From healthcare to finance or retail, the sector-specific nuances significantly impact the choice to Make, Buy, or Ally and as the legislation evolves, industry leaders will have to pay careful attention to the nuances of their own AI landscape. It is very likely that many organisations will default to the Ally decision, in order to share some of the burden and risk of AI integration and use. Already we are seeing different industries developing unique AI solutions with rather different priorities:

• Healthcare: Collaboration between Google and Ascension in AI-driven healthcare data analysis reflects the industry's focus on personalised patient care and regulatory compliance.
• Finance: JPMorgan's in-house AI for risk management shows how financial firms are using AI for decision-making and compliance.
• Retail: Amazon's use of AI for supply chain optimisation highlights the industry's focus on customer-centric solutions.

Understanding industry-specific requirements is a necessary prerequisite for any AI Strategy, but particularly those seeking to utilise alliances and outsourcing to access AI Applications and capabilities. The complexities involved in customising AI solutions should not be underestimated, and leaders should be focused on the need to harmonise any Ally-developed AI with their organisation's governance needs rather than let an external technology partner dictate the play.

The Future of Organisational AI Alignment

As AI continues to advance, businesses must remain agile and forward-thinking in their integration strategies. Emerging trends and technologies will inevitably influence the Make, Buy, or Ally decision, and what works in today’s industry may not work in the future. Frustratingly for leaders, this dynamic can evolve quickly due to the speed at which AI technologies and regulations are developing and changing. Long-term success in AI integration hinges on a proactive and flexible approach, embracing innovation while maintaining a steadfast commitment to ethical considerations, stakeholder engagement, and responsible development. This places an especially high burden on AI Governance systems, compared to other forms of IT or Digital governance we have seen in the past. For leaders to meet this challenge, they will need to begin building strong organisational systems, aligning AI investments with long-term strategy, while also investing significantly in AI knowledge and skill development to support their staff across the organisation. 

Leadership Takeaways:

• Industry leaders Microsoft, OpenAI, Anthropic and Google have formed The Frontier Model Forum to represent their interests in the global AI policy conversation. 
• Companies like AXA are starting to deploy In-House AI on a large scale to harness AI Applications advantages while increase security and governance.
• Social Media companies like Meta are set to experiment with greater chatbot deployment to take advantage of AI developments and enable more personalised marketing. 
• Organisations face a tough decision when it comes to developing new AI tools, between In-House development, external vendors, or strategic alliances.
• The Make, Buy, Ally decision must be taken in the context of both the needs of the organisation’s AI Strategy and with due consideration for the AI Governance Challenges the chosen path will present. 
• SMEs offer the greatest opportunity for the positive impact of AI Applications, though many SME leaders will need help in accessing and deploying the technology. I would call for a proactive policy here to aid SME leaders in this endeavour. 

That’s it for this week’s edition of AI Strategy Brief. Thanks for reading and I hope you enjoyed this update and learned something new. If you found something of value, please subscribe and share this article with others. And if you have any questions, comments, or suggestions, I’d love to hear them in the comments section below. 

See you next time!