Artificial Intelligence, Deepfakes and Chatbots
Deepfakes are a form of manipulated media that use content to make convincing impersonations of voices, images and videos. They can be surprisingly lifelike and are sometimes used to make traditional schemes, like
family emergency fraud and romance scams, more realistic.
Although AI presents
alarming new risks, it also offers many benefits, such as improving productivity and fostering creative thinking. In some cases, AI tools can even help detect fraud. However, be wary of the information you share with AI tools. It could end up in the wrong hands.
How AI Helps Scammers
- Scammers use AI to enhance their fraud schemes.
- With only a few seconds of recorded audio, AI can analyze speech patterns, pitch and accent to create a highly realistic voice clone that can trick close family members or employees into sending money to scammers.
- AI can help scammers create more believable phishing messages or fake documents, making them harder to detect.
- Chatbots are used to conceal scammers’ identities from their victims by communicating fluently in a non-native language.
- Video chats can be manipulated to alter someone’s face and still appear real.
- AI-powered chatbots store and share your data.
- Be cautious of sharing personal identifiable information and copyrighted material with chatbots. Data shared may be used to train and improve the model, and your copyrighted content or personal information may help generate answers for other users. In the event of a security breach, your data could be exposed for bad actors to view.
- If you receive an unexpected call, verify the individual’s identity before responding to any requests.
- Even if the caller ID looks legitimate, listen closely to the audio. Manipulated audio may include choppy sentences, odd phrasing or out-of-place inflection. Use a verified number to directly call the person and confirm whether they attempted to call you.
- If you suspect a video or photo might be a deepfake, look closely at the person’s head.
- Signs of deepfake manipulation include unnatural facial movements, mismatched lip-syncing, unrealistic hair textures or blurriness around the edges of the face.
- Don’t give in to urgent requests.
If you receive a request from someone saying they’re a family member, close friend or co-worker, think twice. Always verify the request before responding to messages expressing urgency or distress.