According to the SEC, Global Predictions Inc. and Delphia (USA) Inc. claimed to their clients that they were using AI in certain exciting ways when – as Gary Gensler, chair of the SEC, concisely put it – ‘they were not’.
This has become known as ‘AI washing’. It describes untrue or deceptive claims about the use of AI that are made or marketed by companies or investment advisers. The companies’ and advisers’ motivation for making such claims is to try and look more sophisticated, innovative or intelligent than they are. AI washing might seem like a minor thing to some (don’t all companies try to put their best foot forward?). But there’s a lot of money to be made by doing it. There’s still so much excitement around AI that exaggerating how much, and how creatively, you’re using it can be very lucrative.
But there’s an assumption that lies behind AI washing that we should scrutinise. The assumption is that the businesses that use AI the most are the best ones. You might think this is true. After all, AI can increase efficiency, bring down costs, come up with ideas and improve decision-making. It can even tell jokes for those times when your work’s wearing you out. But more doesn’t, in fact, equal better in the case of AI. And there are a number of good reasons for this.
First of all, if a task traditionally performed by a human being is outsourced to AI, then the person doing that job suddenly becomes redundant, and is liable to have to leave the company. Some will say that’s the harsh reality of the business world: if someone no longer brings value to the business, then sorry – they’ve got to go. But if it were not true that people bring value beyond what’s written in their contract – value that can be hidden, or might not be quantifiable – the sudden loss of one or more team members always has an impact on morale. When someone loses their job, it’s natural for those around that someone to wonder: ‘Am I next?’ That will affect their ability to work. It might leave some considering their options.
What’s also important to realise is that when we lean on our technology to think for us, our ability to think critically or creatively – a hugely important life skill – starts to atrophy. We get better at thinking well by trying to think well. If we have GenAI come up with all of our ideas for us, or do all our writing for us, then we rapidly lose our ability to do those things ourselves, and with that, some of our ability to make the right decisions in work and life. The people who work in the very highest positions in business, politics and the nonprofit sector are called decision-makers for a reason: the majority of their time is spent deciding. They weren’t born that way: they built up and refined that skill over years by leaning into difficult situations, weighing up all the different options, and making the call. Relying too much on AI for our thinking is a bit like copying someone else’s homework. We lose out in the long run.
The other problem is that the current AI infrastructure is not impregnable, yet the technology is so exciting that basic caution is often cast aside. Something similar happened when tech companies began taking data from their users: users found their technology so convenient that they freely gave away their personal information – yet would never let a stranger rummage through their drawers at home. AI systems are vulnerable to hacking and prone to data breaches, and there have been instances of personal information getting leaked or lost. At the company level, a problem like this can be costly in all sorts of ways.
AI bias can, too. In fact, it’s arguably the most important reason to be cautious about over-using AI. There have been so many instances of this, even at major companies, that the presence of bias in AI is not undeniable. Amazon had to drop a hiring algorithm after finding it favoured candidates who used words like ‘executed’ or ‘captured’ – words more commonly found on men’s CVs. Predictive policing tools have directed more officers to Black and minority neighbourhoods due to biassed data sets. A UNESCO study has revealed hugely regressive gender biases in large language models. From image generation to online advertising and healthcare, bias has been detected time and again in AI, and integrating it too soon or too fully into your company risks entrenching prejudices that people have worked tirelessly for years to eradicate.
And companies should also be mindful that AI has generated a thorny landscape of legislation and regulation that’s evolving rapidly. Lawmakers are debating how to guarantee data protection, promote ethical AI use and respond to industry-specific AI-related challenges promptly. When, in their rush to integrate AI, companies fail to take the time to grapple with these new regulations, they make themselves vulnerable to incurring hefty legal fines.
But truly forward-thinking companies know all this. The truly forward-thinking companies recognise that people, not technologies, are the past, present and future of business. They’re not anti-AI by any means; as much as anyone else (and sometimes more) they understand its enormous potential. But they see its potential to free human beings from menial tasks so they can do what human beings do best: be creative, be strategic and be empathetic. Put more simply: be human.
Michaela Jeffery Morrison is the Founder & Managing Director of the Women In Technology World Series at Techoraco.
