For a moment consider these statistics – over the next decade, Artificial Intelligence (AI) could generate an additional $15.7 trillion worth of additional economic activity across the world, leading to Global GDP increase of as much as 14% (PWC). AI could increase labour productivity by up to 40% by 2035 (Accenture). In contrast, the end of 2019 and the beginning of 2020 have thrown quite some curveballs at corporate leaders across the globe – the trade war between US and China, the uncertainty over Brexit, the escalation of tensions in the Middle East, the uproar about digital interference in elections – to name but a few. In this highly complex and dynamic environment, it takes a lot of time and effort to navigate through the myriad of information and come to a decision as to what is needed and what can & should be done. That is why, and based on the aforementioned numbers, there has been a lot of clamour recently to leverage AI in a better and wholesome way to sift through the almost immeasurable content at hand to separate the wheat from the chaff. Therefore, and unsurprisingly, businesses are falling over each other to become the kings of AI in the new world! Every company claims to be the latest saviour of the overloaded masses by repeating those two famed phrases – automation and AI! The question is really whether these are only empty slogans or is there actually a strategy, and underlying policy, to boot? The answer sadly tends to align to the former. But first, policy makers must understand the difference between automation and AI, which till date the business world has taken for granted by using both terms interchangeably. As the name suggests, automation is all about focusing on streamlining and automating repetitive, instructive and mundane tasks. AI relates to mimicking human intelligence, actions and decisions. In a sequence of events, automation generally comes before AI. If one looks at deep learning, in the form of Bloom’s taxonomy, automation aligns to the first three levels only; i.e. knowledge, comprehension and application. While AI relates to the last three levels as well; i.e. analysis, synthesis and evaluation. And the last three levels are where things get tricky for AI! Analysis includes comparing, contrasting and classifying information to derive a model to understand the information better. Synthesis allows for creating, designing and improving upon the result of the analysis to produce and propose one or a variety of ways forward. Evaluation, the final and most complex rung of Bloom’s taxonomy, includes judging, verifying, assessing the way forward to justify and implement a recommended decision. One can now see why an apt business strategy for AI is crucial! Consider. Evaluation, the final and most complex rung of Bloom’s taxonomy, includes judging, verifying, assessing the way forward to justify and implement a recommended decision. One can now see why an apt business strategy for AI is crucial! Consider Previous op-eds have defined policy as guidance that is directive or instructive; i.e. it is clear in stating its objectives and what is to be accomplished. It is a galvanizing vision that describes the end goal. Strategy is defined as ‘ways and means to an end’. That end being the defined objectives established by the policy. In its entirety, strategy is a continuous process where ends, ways, and means are aligned to accomplish desired policy end goals while keeping risk at an acceptable level. Therefore, AI at best is a strategy that can achieve the end goals described by the overarching policy objectives. While policy is to be decided by the echelon of corporate leadership and it will definitely vary from company to company, an appropriate policy statement – that uses AI as an executable strategy – is, transformation of ways of working to repurpose humanity for higher level thinking! If this policy could be expressed in one word, it would be this: ‘repurpose’. For this overarching policy of repurposing humanity, the following five-prong approach should be used to elaborate an AI corporate strategy. One, ensure that the AI corporate strategy aligns with the overall business strategy which – in turn – will stem from the overarching company policy. Note that this linkage is crucial to assure success of this whole endeavour. Two, highlight the correct problems to solve with AI. These can be divided into tactical AI priorities and strategic AI priorities. The former providing the quickest & biggest bang for the buck and the latter envisaging a return only in the near future. Three, because almost all successful AI initiatives bank on the rationalisation and understanding of tons of data, understanding of big data and a complimentary data strategy is critical especially in light of the chosen tactical and strategic AI priorities. Four, establish a dedicated cross-domain team to assess and highlight the gap in skills and capacity which may act as a dependency for the implementation of the selected AI priorities and leverage and enhance the data strategy as needed. Five, discuss and finalise an approach to manage change and related ethical and legal issues that will arise because of the use of AI. All of this should be captured in a formal AI strategy document which then can easily underpin not only the business strategy but also the related policy objectives. In many ways, handling of AI within the corporate environment is a poisoned chalice! But it can turn out trumps for business leadership if handled well; i.e. a realistic corporate AI strategy that caters for the aforementioned approach, underwrites the business strategy and aids the realisation of policy end goals. Anything less would be akin to falling over one’s own sword! The writer is Director Programmes for an international ICT organization based in the UK