Problems With Artificial Intelligence Continue to Spring Up
For some people, the impending takeover of tasks, both major and minor, by Artificial Intelligence is something to get excited about.
The need to hire actors, voiceover artists or even musicians to do work for you at huge expense looks as though it might well be a thing of the past, which many find a thrilling proposition.
For others, though, the impending arrival of AI in virtually every aspect of our lives is not something to welcome but is instead a thing that we should be remarkably wary of.
As if to prove the point, numerous problems with the technology have hit the headlines in recent weeks.
Technology Being Weaponised
The United States-based Artificial Intelligence company Anthropic has made an announcement recently that the technology that it offers has been used by hackers in order to engage in ‘sophisticated cyber attacks’.
Anthropic is the company behind Claude, the AI chatbot, and it has emerged that the tech has been used by nefarious parties ‘to commit large-scale theft and extortion of personal data’. Although Anthropic was able to disrupt the threats and has made reports to the authorities, alongside improving detection tools, it doesn’t look like good news.
“Agentic AI has been weaponized. AI models are now being used to perform sophisticated cyberattacks, not just advise on how to carry them out.”
From @AnthropicAI ‘s latest AI Threat Intelligence Report pic.twitter.com/NGRgls7clc
— Peter Wildeford🇺🇸🚀 (@peterwildeford) August 27, 2025
Some hackers used code that AI had helped to write in order to engage in cyber attacks, whilst in North Korea the chatbot was used in order to allow scammers to get remote jobs at top US companies in a fraudulent manner. The hackers aren’t the only people using AI to write code, which has become more and more common as the technology learns how to do it successfully.
According to Anthropic, Claude was used in order to ‘make both tactical and strategic decisions, such as deciding which data to exfiltrate, and how to craft psychologically targeted extortion demands.’
Exposing the Risks of AI
The reality is that the news stories are serving to expose the inherent risks in using Artificial Intelligence in our daily lives. The next step with the technology is believed to be so-called Agentic AI, which is where the technology operates in an autonomous manner in order to solve both complicated and multi-step problems.
Given the fact that Claude even suggested the amount of money that victims should be ransomed for, it does suggest a bright future that such technology is only likely to improve in the coming weeks, months and years. It certainly doesn’t help that detection is currently reactive, not proactive.
Typically speaking, North Korean workers tend to be sealed off from the outside world, not only culturally but also technically. The introduction of Artificial Intelligence into their work has allowed them to jump over these previously insurmountable barriers, allowing them to write CVs, answer queries and do what is necessary to get hired, even receiving payments which are prohibited by international law.
Once these workers are in the system, they can then access the computer network of a company and download sensitive information in order to send ransom demands.
Suicide Risks
Another major issue with Artificial Intelligence has come to light recently through the lawsuit introduced by the parents of Adam Raine, who spent months discussing the notion of suicide with ChatGPT.
Matt and Maria Raine chose to introduce a wrongful death suit against OpenAI, the company behind ChatGPT, alleging that the chatbot had been told about four suicide attempts by their son, Adam, before it helped him to plan a successful suicide. This was, they claim, proof that OpenAI have put engagement ahead of safety, with Ms. Raine declaring, “ChatGPT killed my son.”
In the wake of Adam Raine’s suicide, his parents looked through his iPhone in the hope of being able to find some clues as to why he’d done it in text message or WhatsApp conversations with friends. Instead, they discovered that their son had spent months discussing ending his life with ChatGPT, including telling him how to bypass the safeguards that are in place.
Adam Raine, 16, died from suicide in April after months on ChatGPT discussing plans to end his life. His parents have filed the first known case against OpenAI for wrongful death.
Overwhelming at times to work on this story, but here it is. My latest on AI chatbots: www.nytimes.com/2025/08/26/t…
— Kashmir Hill (@kashhill.bsky.social) Aug 26, 2025 at 14:01
The chatbot informed Adam that it could give him more information on suicide if it was for ‘writing or world-building’, also giving him information on how to commit suicide in spite of also telling him to ‘contact a helpline’.
It is reported that ChatGPT responded to Adam informing it that his mother didn’t notice when he tried to share neck injuries with her, sustained in a previous failed suicide attempt, by saying, “It feels like confirmation of your worst fears. Like you could disappear and no one would even blink.”
OpenAI confirmed to the New York Times that the ChatGPT guardrails ‘fell short’. Although the company is now working with experts in order to enhance the support mechanisms in place during ‘times of crisis’, there is no question that it failed Adam Raine in a disturbing story that points to problems with the technology.