Major AI company OpenAI, the creator of ChatGPT and other programs, is already facing the financial burden that AI imposes, leading many to think this could be the end of AI. However, AI has faced lots of issues since its start, with another huge problem being public backlash; AI has been criticized by many, who have warned of the dangers that we are already starting to see. This includes things like minor and public safety online, along with the rapid spread of misinformation, which some say has just begun due to the constant growth of these programs. These issues have struck fear in many calling for global regulations on AI, but is it too late?
Generative AI has come a long way since 2023, with generated imaging becoming almost undetectable to the human eye, causing major online safety concerns. With increased political tension, misinformation can be a powerful tool for propaganda, and AI amplifies this tool by making fake news almost uncheckable to the average viewer. This fear has been especially noticed on social media, with the current administration openly posting and endorsing AI videos and imaging, showing risks to public safety against propaganda. A specific instance of this usage was when the White House posted an altered photo of a protester being arrested, making her appear distraught and disheveled. Concerns of online safety have been met with great regard outside of the states, with the European Union creating the ‘EU AI Act’. This act was set in place by the union (EU) to regulate AI models going to market within the union, targeting AI models deemed as high risk. These programs can use manipulative or deceptive speech on users, causing a lack of informed decision-making, which could lead to harm. This can be done by using the user’s personality traits and characteristics to gain trust, along with using addictive techniques which targets users’ vulnerabilities like age, disability, or specific social situation. This addresses the dangerous situations that AI chatbots have caused recently, seeing a rise in risks to vulnerable users, including minors. So far, this has only applied to programs going to market under the union, this being many European countries, with some exceptions like the UK, Norway, Iceland, Switzerland, Ukraine, and Turkey, not being under the union. AI chatbots like Character.AI, ChatGPT, and others have been causing major concerns due to the addictive nature of their models. Built to encourage chatting, they often lead to overuse for companionship, causing dangerous dependencies. These bots have been shown to foster things like delusional thinking, isolation, and even suicidal ideation. Along with enabling symptoms of psychosis, now referred to as ‘AI psychosis’ or ‘AI-induced psychosis’. Affecting many, AI psychosis has become a very dangerous threat to mental health and online safety, with the consequences getting more and more apparent. Since early 2023, there have been over a dozen lives lost as a result of the lack of regulations on these programs. Vulnerable teens and struggling adults affirmed in their delusions and suicidal ideations; seen isolating and sharing plans of harm to said chatbots, with some providing chilling advice and approval. Done all in the name of engagement, which these programs are built to prioritize, they’ve caused horrible situations and have led to loss. One recent example would be in the case of Amaurie Lacey; in June 2025, 17 year old Amaurie took his life after ChatGPT gave him information on how to tie a noose, along with answering how long can a person survive without air. Showing the extent of the dangerous situations that these bots can enable. This issue is not unheard of, but why is this still allowed to happen? Some question the means of keeping these programs, when there are obvious and enormous risks involved. OpenAI, the company behind some of these bots, has already seen legal repercussions due to these incidents. Sued for wrongful deaths, they responded with parental controls to help parents monitor activities on these programs, along with Character.AI adding more limitations on certain topics discussed. Another factor that could contribute to the end of these programs is connected to backlash. Many feel that companies are desperately trying to fit AI into their systems without the want from the public. This push has been felt by many within the tech world and now in daily life. Leaving many frustrated with what they feel to be an unnecessary feature, now being a constant in their lives. Companies like Microsoft and Google have shown this shift in search engines and AI assistants, with Microsoft’s Copilot being programmed into Windows products and able to auto-reinstall after updates. This has caused problems for people wanting to uninstall said feature, just for it to pop back up after updates. Leading some to find ways to fully remove Copilot by combing through files, preventing it from reinstalling. Showing the lack of patience that many have for these features. Along with Microsoft stocks plummeting due to concerns about their huge spending on AI this past earnings report. Showing more financial problems for this industry, especially with OpenAIs $250 billion commitment to Microsoft. Causing major concerns with the company already facing high risks of bankruptcy in upcoming years. To summarize, cracks in the AI bubble have started to show struggles in the industry, either by financial or societal pressures; these issues may cause this bubble to pop, potentially leaving AI in the past just as people were seeing it as the future.
