Why companies should have an AI business strategy in place

Companies across many industries are increasingly becoming AI-charged, which means they’re using the technology to drastically increase productivity and efficiency across a range of specific use cases. However, the rapidly rising number of AI applications requires companies to develop a coherent and comprehensive business strategy around the technology. 

AI will soon be integrated across products, services, operations, support, and governance which means companies will need to actively break down internal silos to avoid duplicated efforts and misalignments. Another critical priority during the AI revolution is talent acquisition and retention. While powerhouses like Google, Meta, and Amazon have monopolised a significant share of talent in the field, there has also been a profusion of startups using public data to create new products and new jobs. Employees recognise that the cultivation of skills relevant to developing, implementing, and working with AI will only become more important in the coming years, so companies need to be on the lookout for talent in many places. 

AI will fundamentally alter the business landscape in the coming years, and companies need to be prepared for this shift. This means developing internal processes to implement and manage AI applications while orienting their business strategies toward an AI-charged future. 

Overcoming AI integration challenges

AI adoption rates were already steadily rising before the mania around generative AI that began at the end of 2022. According to McKinsey, 40% of companies say they will increase their overall investments in the technology due to breakthroughs in generative AI – a proportion that will likely increase. Accenture recently reported that just one type of generative AI, large language models (LLMs), will have an impact on 40% of working hours. 

The mass adoption of AI makes it all the more important that companies understand how to integrate the technology with their workforce and processes. This could mean abandoning clunky and inefficient legacy systems, improving data security and privacy, and developing a more centralised digital infrastructure. IBM found that two of the top obstacles to AI adoption are projects that are “too complex or difficult to integrate and scale” and “too much data complexity”. AI is only as good as its underlying data, which presents a problem for enterprises that have large amounts of noisy and fragmented datasets. 

Companies know they will eventually need to move beyond its use for individual point solutions, as independently designed use cases can lead to the large-scale misalignment of goals, unnecessary complexity and operating costs, and redundant processes. To fully leverage AI, companies will have to develop a holistic strategy for implementation and operation. 

Evaluating and addressing risk

Within two months of ChatGPT’s launch in November 2022, there were 100 million monthly active users of the platform. Despite the widespread enthusiasm for AI, the limitations of the technology have also come into sharper focus over the past year. For example, LLMs have a tendency to “hallucinate” – they present false information in the same compelling way that they present facts, which makes it difficult for users to tell the difference. It also presents serious privacy and cybersecurity issues that many companies aren’t prepared to handle. 

A 2023 survey found that just 38% of companies are mitigating cybersecurity risks associated with AI, only 25% are addressing intellectual property infringement, and 28% are working on regulatory compliance. These are all alarming indicators that companies aren’t prepared for the AI risks they confront, even as adoption rates surge. Companies need an enterprise-wide view of the data their AI models are accessing, as well as any potential vulnerabilities that can be exploited by cybercriminals. At a time when hackers are increasingly exploring how to use AI to launch cyberattacks and regulatory scrutiny is only becoming stricter, it’s essential to proactively address risks. 

An ongoing challenge of its adoption, integration, and management is the technology’s lack of transparency. LLMs, for instance, are black boxes that don’t reveal how they synthesise data and produce content. Beyond the implications for data privacy and integrity, this lack of transparency can make it difficult for companies to ensure that AI models are predictable and usable. Addressing these issues needs to be a core part of companies’ AI adoption strategies in the coming years. 

Developing and implementing an AI business strategy

AI is a revolutionary technology that will radically transform how many companies operate, the products and services they provide, and the ways they engage with customers. This is why companies can’t simply slot AI into their technology strategy. They must build an entire business strategy around AI, which encompasses workforce development, internal and customer-facing processes, and the identification of revenue opportunities. Meanwhile, they will need to use strong privacy and security guardrails to minimise the liabilities associated with the technology. 

Companies should determine what they want to accomplish with AI and construct a strategy around their goals. This strategy will depend on many variables (company size, industry, etc.), and it may have to be altered as market conditions change and the technology evolves. Companies will soon be able to choose from a suite of AI models with multiple use cases that can be deployed depending on their specific needs. This will require data orchestrators who can bring the outputs of these models together in a cohesive whole. 

A 2022 IBM survey found that the top barrier to AI adoption is limited skills and expertise – a powerful reminder that companies need to substantially increase their AI talent pool. Beyond recruiting new employees with requisite skills, companies can also invest in workplace training and professional development. More employees report that they believe AI will help them improve productivity, learn new skills, and pursue new opportunities than affect them in negative ways, and companies should take advantage of this. 

Companies won’t be able to fully embrace AI without the right workforce, digital infrastructure, and strategy in place. As the technology becomes increasingly necessary to remain competitive, companies will need to figure out how to adopt and deploy it as effectively as possible.

Steve Schmidt is a general partner at Telstra Ventures.