“An AI tool is only as ethical as the developers who created it and the source material on which it was trained. You, the human in the know, are pioneering the ethical use of AI.
I was scrolling through my LinkedIn feed recently and noticed that a former associate had posted that he had earned a certification in “AI ethics” from one of the largest tech companies in the world. I’ve noticed that this term has become more ubiquitous lately, and it’s confusing.
Ethics according to whom? Ethics in relation to what? What ethical code do we use to determine whether a given technology is ethical? By what standards do we measure whether an image, song, article, think piece, or other AI-generated asset is “ethical”?
There is a simple answer: ethical AI is a fad. This may look great in a press release, but the reality is that it lacks substance. It is nothing more than a marketing term designed to make buyers feel safe investing in this technology. And why, you might ask, do companies – particularly financial institutions in our case – that seek to enable the use of AI feel unsafe or reluctant when it comes to investing in AI solutions? There are a few reasons.
Trust Killers
Of course, AI technology is nascent and evolving daily, leaving security teams with little time to examine the vulnerabilities of these tools. But ultimately, regulated industries are hesitant to use AI solutions because they are seen as black boxes with limited transparency, making regulatory compliance validation difficult and time-consuming. When it comes to developing and implementing ethical Gen AI solutions, companies at the forefront of this new generation of AI technologies have proven unable and unwilling to effectively self-regulate and ensure transparency that regulated sectors need.
Without transparency, there can be no accountability, and without accountability, guarantees of ethical use of AI are meaningless. This is where sensible regulation can immediately make a difference.
Without government regulation (such as European AI lawfor example) which aim to provide guidance, safeguards and accountability to ensure that consumers are not impacted by technology companies that release hastily trained and poorly tested AI models into production, we receive a constant flow of information that looks like this:
Headlines like these kill trust.
Large companies like Google, Microsoft, Meta, and OpenAI are driven by their desire and ambition to dominate the market in the growing, trillion-dollar AI industry. This zeal often leads them to take shortcuts in their development processes. As a result, when these giants fail to deliver on their lofty AI promises, it’s not just their reputations that suffer. Hardest hit are startups, small businesses, and emerging technology companies that invest in developing truly intelligent, safe, and ethically trained AI products. These responsible entities must deal with the considerable reputational damage caused by large company bankruptcies, which often results in more frequent, costly and embarrassing headlines.
Ethical AI is OUR Job
I’ve worked in tech long enough to understand hype cycles, and we’re certainly in the middle of one with the AI generation. With all businesses clamoring for attention in the new world of AI, terms like “ethical AI” and “responsible AI” play on fear rather than value. “You’re afraid of AI,” these messages seem to scream, “but if you use our product, you don’t need to be afraid.” Tech marketers should leave the scaremongering to politicians and stick to the facts.
The ethics of AI as a service is bananas. If a company promises “ethical AI,” run away. An AI tool is only as ethical as the developers who created it and the source material it was trained on. You, the human in the know, are pioneering the ethical use of AI.
So when you’re considering partnering with a company offering something akin to “ethical” or “responsible” AI, remember that it’s your job. Rather than seeking certification from one of those companies that can’t be trusted to self-regulate, perhaps spend this time delving deeper into the term “irony.”
Image source: photo repository
Image ID: 6496641
Copyright:: Stuartmiles
