Delivering business value through AI

AI is being used by every industry to achieve greater efficiencies, deliver more effective and affordable services, and enable better customer experiences. There are thousands of examples of the success and growth it has enabled. 

Across all industries ranging from healthcare, education, government, charities, and telecommunications to media, retail, travel and tourism, transportation and logistics, and manufacturing, AI has been embedded into processes, systems, and procedures to add genuine business value. 

Caveat emptor

The positive impact AI has and will continue to have on businesses, society and government is well understood. However, alongside the positives there is a stream of negative headlines voicing concerns about AI. Words and phrases used to preface AI, such as ‘out of control’, ‘deepfakes’’, ‘misinformation’, and ‘manipulation’, to name a few, are eroding trust. 

This has been a wake-up call to organizations. They are now well aware that whilst the benefits of AI are numerous, without the correct human oversight, control, and partnership with an experienced service provider with the required technical, knowledge management, and security, compliance and risk expertise, they will be open to the possibility of serious reputational damage, and an erosion in brand trust. 

Humans allow AI to be the instrument that damages

At the start of this month (January 2025), AI was once more the subject of a negative headline. Apple’s AI feature that summarizes news articles, launched with much fanfare in October last year (2024), as Apple Intelligence, came in for serious criticism and calls for market withdrawal after a series of errors. A few weeks later, (January 17^th) Apple took the decision, after calls from news outlets and journalist unions, to remove the feature. Humans released it and humans rescinded it. 

AI is not ‘out of control’ it just hasn’t been implemented with controls

Using generative AI to summarize breaking news, and BBC content as the source, Apple Intelligence was accused of spreading highly inaccurate information. The headlines generated by Apple Intelligence were very plainly wrong, causing reputational damage to the BBC and to the journalist who had written the original, and correct news item. Apple’s news summarization feature seems to be reimagining some news articles, turning them into misinformation, and sharing them to appear as source text from the original outlets.

The BBC is not the only news outlet to have had its news misinterpreted by Apple Intelligence. This has caused, as the organisations themselves have said, “significant reputational damage”. Apple Intelligence sent alerts, summarising The New York Times, that spread false claims about arrests (that didn’t happen), and inaccurate summaries about events. 

Sky News and The Washington Post have also been named as organizations that have had their correctly reported news summarized into headlines by Apple Intelligence that are misleading and inaccurate. The BBC has issued a formal complaint to Apple, citing instances of Apple Intelligence generated headline summaries, ranging from claims a man had shot himself (he hadn’t) to the winner of a sports event before it had taken place. 

In-control of AI 

Technologically advanced, sophisticated, context aware, inbuilt guardrails, controllable and verifiable knowledge management, secure, compliant, and governed. These are the standard features of a modern conversational AI platform that enables ‘in-control AI’. 

Implementing AI as a strategic tool requires experience, expertise and understanding of the intricacies and complexity associated with large volumes of content from multiple sources and applications, knowledge management curation for contextual understanding, automated accuracy, due diligence and local and regional compliance. 

The reputational damage stemming from misinformation, information that misleads customers, incorrect information, and fake information can be avoided. When the correct controls, protocols and procedures are automated through a conversational AI platform that acts as the singular management and orchestration center for an organization, summarization can be done correctly, and the output will be accurate. 

Avoiding rogue, and wrong, contextual interpretation is possible when humans are in control of AI. AI can and should be ‘domesticated’ for purpose. It should not be used, without oversight, in the virtually ungovernable universe of the online world. 

V-Studio

AI is here to stay and will forever more be part and parcel of content generation and sharing. This puts the moral and professional onus on organizations sharing this information to be in control and have oversight and ownership of the content shared on its behalf. 

Trust between brands and their customers is only possible with accuracy, integrity, and consistency when it comes to information, conversations, and engagement. Creative Virtual’s award-winning V-Studio platform, gives organization control over the AI they use. 

Human control over AI ensures they are managing and are fully accountable for their reputation. Each organization decides for themselves the risk appetite, compliance standards, and governance protocols to put in place around the use of generative AI and curated content that will safeguard their brand, their employees, and their customers. 

AI needs human intelligence

AI should not be seen as a ‘bandwagon’ to be jumped on. Implementing it requires human intelligence and expertise. It will not be the great panacea to solve all business ills if it is rushed and implemented with a sticky tape and staple approach. 

The flaws that threw up the incorrect Apple Intelligence news summarizations and forced organizations to call for its withdrawal are avoidable. AI, like humans, are fallible. But with humans in control, AI fallibility is reduced. 

Every organization, whatever the industry, needs to seek out experienced, expert AI service providers who know more than just the technology. Unless organizations look to companies that have a multi-disciplinary approach to conversational AI – technology, knowledge management, risk, security, and compliance – AI will continue to be in the headlines for what has gone wrong. But there are many more rights than wrongs. 

Organizations that want to fight the good fight do it with humans, and take a human approach to AI.