The dangers of AI in the workplace

Home » News & Blog » The dangers of AI in the workplace

Artificial intelligence (AI) and generative pre-trained transformers (GPTs) have changed the way businesses operate, automating many processes and improving the efficiency of mundane, everyday tasks. However, with the advancement of AI technology, new risks for businesses have emerged – particularly from deepfakes.

These AI-generated forgeries can deceive employees to leak data and unwittingly transfer funds, manipulate decision-making and even damage a company’s reputation. Discover what deepfakes are, how to spot them and how businesses can protect themselves from the threats of AI.

Deepfakes – what are they?

Deepfakes are forged videos, images and phone calls created using the power of AI, where a person’s likeness and/or voice is replicated to make them appear to say or do something they never did. These forgeries can be incredibly convincing, making it difficult to distinguish between real and fake content.

In the workplace, deepfakes pose serious risks, such as:

• Fraudulent transactions: A scammer could impersonate a CEO or senior employee in a video or voice call, pressuring employees to transfer money or share sensitive data
• Reputation damage: Fake videos or statements attributed to company leaders could spread misinformation
• HR issues: Employees might create deepfakes of colleagues which results in harassment or defamation

How to spot a deepfake

While deepfakes are becoming more sophisticated, luckily there are still ways to detect them. Here’s what to look out for with a suspected deepfake:

Unnatural movements or inconsistent speech: Look for odd facial expressions, blinking patterns or lip-syncing that doesn’t quite match up

Audio inconsistencies: AI-generated voices may sound slightly robotic, have unnatural pauses and cadence, and follow unusual patterns

Poor lighting or shadows: Deepfake videos sometimes have inconsistent lighting or blurry edges around faces

Verify the source: If you receive an unusual or unexpected request via video or audio, confirm it with the person through another communication channel (e.g. a phone call or email)

Trust your instincts: If you feel like something is suspicious about a video or voice call, trust your gut and question it. Scammers using deepfakes will often make their targets feel like they must act immediately. Taking your time and scrutinising the video or voice call received can help determine whether it’s a deepfake

Protecting your business from the risks of AI

To protect your business from the dangers of deepfakes, you should take proactive steps such as:

• Employee training: Educate staff on the risks of deepfakes and how to recognise them
• Two-step verification: Require secondary confirmation for sensitive requests (such as financial transactions)
• AI detection tools: Invest in software that identifies deepfakes
• Clear company policies: Establish guidelines on how to handle suspicious communications and the repercussions of falling for a deepfake scam

Can an employee face disciplinary action if they fall for a deepfake scam?

If an employee unknowingly falls victim to a deepfake scam, such as transferring funds based on false instructions, disciplinary action can depend on the circumstances and severity. These include:

• Whether the employee followed company policy
• If it was negligence or a lack of training that contributed to the mistake
• The severity of the financial or reputational damage

In most cases, businesses should focus on improving employee training rather than punishing them, as deepfakes can fool even the most cautious individuals.

Have deepfakes had any high profile success?

A multinational engineering firm was defrauded $25 million (approx. £19.8million) after a finance worker took part in a video call with what they thought were senior colleagues.

They were instructed to make several bank transfers in the video call, which they did willingly. All the involved colleagues in the call turned out to be deepfakes.

The employee was initially suspicious about the request to join a video call, but because their colleagues looked and sounded like those they recognised, they didn’t question it further. It took the company more than a week to realise that substantial fraud had taken place.

What happens if an employee creates a deepfake image?

Creating a deepfake of a colleague, whether intended to be malicious or not, can have serious consequences, including:

Disciplinary action: The employee could face suspension or have their contract terminated

Legal repercussions: Depending on the use, deepfakes may violate harassment, defamation or privacy laws. This can lead to fines and even prison sentences in some circumstances

Reputation damage: The company’s image could suffer if it’s associated with unethical AI use

Businesses should have clear policies prohibiting the malicious use of AI and outline the consequences for any violations.

While AI is a useful tool when in the right hands, businesses should be aware of the risks associated with this emerging technology. If you’re worried about the risks involved with using AI or are concerned about the threat that deepfakes pose to your company, get in touch with one of our employment law experts for professional advice.