Deloitte Australia has issued a formal apology for its A$440,000 report. The report, commissioned by the federal government, was riddled with mistakes generated by artificial intelligence (AI). This event has raised significant issues around AI trustworthiness throughout the workforce. A recent survey indicates the majority of employees unwilling to be transparent about their AI tool use.
This was the conclusion of a recent representative study conducted by BetterUp Labs and Stanford Social Media Lab. They discovered that over 61% of workers say they don’t disclose when they use AI to do their jobs. What we learned According to the survey, over half (55%) of employees had taken credit for AI-generated material. This development leads to some critical ethical issues regarding workplace transparency and accountability.
This is yet another instance that shows us the troubling, increasing reliance on AI with little oversight or training.
AI training & guidance
Less than 50% of surveyed employees indicated they had received any kind of training or policy guidance on using AI tools. The failure to adequately educate has fostered an attitude of complacency that extends to accepting AI outputs at face value. Yet, shockingly, nearly two-thirds of employees confess that they use AI-generated content without verifying its accuracy.
The Hidden Burden of AI Dependency
The implications of this unrecognized reliance on AI tools are starting to play out – and the pattern isn’t good. More than one in four employees said AI has made their work more demanding. This new layer of pressure makes their routine tasks take even more time. This unintended consequence is compounded when you consider that fixing problems with AI-generated work tends to take roughly two hours per use. That’s a lot of time wasted on fixing things!
For the big companies, the fiscal consequences are shocking. The average productivity related AI failure for a company with 10,000 employees is estimated to cost that company US$9 million per year. That’s about A$13.8 million in annual foregone revenue from dotage. These statistics highlight the importance for businesses and nonprofits alike to mitigate the negative impacts of bad AI usage.
Employees’ unchecked reliance on AI can severely impact productivity, but more importantly, undermines producing quality work. Our survey revealed that critical engagement with AI—like meaningfully evaluating its outputs—is almost nonexistent. It’s up to employees to critically assess the output produced by these systems. If they fail to do so, they will inevitably be introducing fatal flaws into their work, similar to the flaws in Deloitte’s report.
The Importance of Training and Policy Guidance
Evidence clearly shows that increasing AI literacy via training is a prerequisite for engaging productively and critically with the technology and for mitigating adverse mistakes. Employers need to laser-focus on offering comprehensive learning opportunities and detailed instructions on the acceptable use of AI technologies.
Yet today, many of these same employees report feeling woefully unprepared to deal with the high stakes nature of AI technology. Under a third say they have received any training or policy direction on AI. This leaves a steep learning curve that organizations need to overcome. AI literacy proper training can help employees use AI to its full potential and make sure employees critically engage with its outputs.
Furthermore, taking a proactive approach to AI presents a huge opportunity for positive impact. Over two-thirds of employees say working with AI tools has made them more efficient and generated a sense of innovation in their job. But without the right training and oversight, we’ll never realize these potential benefits.
The Need for Cultural Change in AI Adoption
The results should be a wake-up call to the administration and lawmakers that we must rapidly change our culture surrounding AI adoption in the workplace. Employees often conceal their use of AI, indicating a lack of supportive policies and an atmosphere where transparency is not prioritized. Businesses need to create a culture where employees feel safe discussing how we should or shouldn’t use technology.
It’s through these big changes to company cultures that companies can push toward implementing overarching policies on what AI can and cannot do, encouraging ethical AI use. Further, by promoting a culture of learning, you will empower your employees to ask questions and seek help in understanding how AI can be applied.