Know the dangers of AI
While not directly involved with vessels, the reliance on AI in the workplace and in general, is alarming.
Have your heard of AI Poisoning?
AI poisoning occurs when attackers post false information online to train AI. For example, they might create fake customer service numbers or write misleading comments. AI tools can then repeat this false information as if it were true. Every AI search or chatbot is at risk of this attack.
While AI responses often look professional and sound confident, they can be misleading. Because these models don’t automatically verify source credibility, they can be manipulated to display inaccurate information. Always verify data sources when searching for sensitive information, like customer service numbers, login pages, or payment portals.
AI responses are summaries generated by AI from information gathered from websites, forums, review sites, and public comments. These sources are vulnerable to AI Poisoning. While these summaries save time, always verify the sources, as they do not guarantee that information is correct or safe.
Scammers use AI poisoning to commit financial fraud and credential theft. They may direct victims to scam phone numbers, request payments or refunds, or steal credentials by sending victims to fake login pages and capturing login information. By having AI repeat their information, they make their scams appear more legitimate, increasing their chances of success. Dangerous, isn’t it?
Use AI responses as a starting point, not a conclusion. They can help you find low-risk, non-sensitive information, such as definitions or explanations, general background information, and comparisons of features or concepts. Do not use AI responses for customer service contact details, login or account access, payment, refunds, or security issues.
Artificial intelligence(AI) is great at finding information, but it’s not great at evaluating who posted it, why it exists, or whether it’s trustworthy. While AI is a more intelligent search engine, the susceptibility to poisoning means it’s not necessarily safer.
Check your sources always, every time. Whilst AI has its many advantages, it can also be misleading, inaccurate as AI relies on the input on the internet. AI does not verify the information it summarises to you, therefore, you may be making your systems/phone etc vulnerable to an attack if not cautious.
AI responses may cite sources that are unreliable, like social media posts, rather than official websites. Before actioning anything for e.g. calling a number, click on a link from an AI response, verify it by visiting the official website.
