Like it or not, AI is here to stay — and will continue to grow in the workplace.
More than a third of companies use AI in their businesses already, and another 42% report plan to use AI in the workplace, according to an IBM study.
So — just like in the advent of social media — companies will need to set the standard for AI etiquette and content.
But curiosity has been known to kill the cat. And, in the workplace, HR and other executive leaders don’t want to employees to take risks with AI that could jeopardize security or reputation (or nine lives!).
Enter AI etiquette
That’s where AI etiquette comes in.
If you’re wondering if that’s important already — after all, it took some time before companies adopted social media policies — it is. Here’s why: Most companies that are already using AI aren’t doing anything to ensure that it’s trustworthy and responsible. In the IBM study:
- 74% don’t check that AI reduces bias
- 68% don’t track its performance, and
- 61% aren’t making sure they can explain AI-powered decisions.
“As with every new technology discovered in the history of mankind, the possibilities of AI are both a curse and a blessing,” says Bergen. “And as with every new technology, its ultimate value to humanity will be determined by how cleverly, thoughtfully, and efficiently we use it.”
To get ahead of AI issues, consider these emerging policies and tips for content and etiquette:
1. Check quality/accuracy
AI may seem like a powerful tool, capable of automating everything from production to filing, and fixing everything from DEI to the bottom line. After all, it can spew out in milliseconds what would take a day for a group of employees to produce. But the accuracy and quality of the information isn’t full-proof. And probably never will be.
That’s why any organization will want to make fact-checking and quality assurance critical elements in the use of AI.
“We check the results of AI tools constantly and adjust them where necessary,” Bergen explains about getAbstract’s approach. “Experience so far has shown us that in many areas, AI tools and their integration do not yet deliver the results that meet our quality requirements.”
Establish a protocol to check AI-generated information for accuracy and quality. You may need a two-party authentication process to ensure the data is correct and well-presented.
2. Require training
Just because employees know how to use computers and access AI tools doesn’t mean they know how to use them properly. That’s probably why 35% of companies are already training or re-skilling employees to work with AI and automation software and tools, the IBM study found.
Nearly anyone can access AI — and sometimes use it to do their work without the boss knowing. So, at the very least, you’ll want to institute baseline training that includes:
- identifying reliable sources
- requirements and guidance on how to check for accuracy
- the degree AI can be integrated in a project and/or role, and
- security/privacy practices.
3. Recognize/address security issues
AI technology can increase efficiency and productivity as it often helps with content creation, data analysis and automation. But AI tools can also pose security risks with serious implications — including data breaches, privacy issues and copyright infringement.
Employees need to understand these – plus the emerging legal implications that are arising with AI. Here are four major risks:
- Data breach. If your systems aren’t covered by cybersecurity, antivirus and file-sharing securities, employees can fall prey to malicious attacks for data while accessing AI tools.
- Model poisoning. This happens when malicious data infiltrates an AI system employees are using. It can corrupt the system and spew out bad or malicious results.
- Plagiarism. AI usually doesn’t produce content that is 100% unique. That means, employees’ searches will be chocked full of someone else’s work, which can be flagged (embarrassingly) as plagiarism.
- Copyright infringement. Similarly, AI can produce images, audio and video that is copyrighted — and using it without proper attribution can expose your company to copyright infringement issues.
4. Address appropriateness
It should go without saying, but all too often, if it’s not said, it’s not avoided.
So let employees know that asking AI inappropriate or offensive questions, or asking AI to perform illegal activities — and sharing information from the answers — is not permitted. This likely already falls under your existing online etiquette policies, but it’s worth repeating.
5. Save space to practice
Encourage employees to try and experiment with AI without consequences — i.e., don’t make their foray part of a project or anything that would be reviewed for meaningful evaluation. Also, recognize that some employees may not want to touch AI.
“We let our employees test in safe spaces, share their learnings, and only offer well-tested and secure solutions to our customers,” says Bergen. “This requires ensuring that they remain curious, are open to new ways of working, and are willing to adopt a trial-and-error mindset.”
6. Be transparent
If you’re using AI for business purposes, it likely affects all employees, plus clients and customers. Make sure stakeholders know that you’re using AI to produce results.
Even better, explain why you use AI, how you ensure it’s accurate and relevant, and what you plan to do with the information that’s been collected.
“Nothing is more annoying than unsupervised AI chat or phone bots pretending to be ‘Chris’ from customer service, stealing your time with unnecessary questions only to put you on hold for human support,” says Bergen, as an example. “That’s the kind of experience we want to spare everyone.”
7. Balance use
AI use in the workplace has the potential to be harmful and helpful.
“AI can help us make faster, more creative – and better – connections between individual bits of knowledge that are structurally related but may not seem connected at first glance,” says Bergen. “If we allow ourselves to blindly rely on AI to tell us what is right or wrong, we will be in danger of creating an idiocracy for ourselves. We must remain critical thinkers and retain our common sense.”
Every organization, department and employee will likely need to find the proper balance of artificial and human intelligence. Help employees recognize that they will have the upper hand in creativity, critical thinking and emotional intelligence. They can work in collaboration with AI on algorithms, automations and probabilities.
8. Evaluate use
AI tools and their application and practicality in the workplace will constantly evolve. That makes it critical for organizations to regularly evaluate its role, uses and etiquette in your workplace.