How Safe Is Artificial Intelligence

AI and deep learning models can be difficult to understand, even for those who work directly with the technology. This leads to a lack of transparency for how and why AI comes to its conclusions, creating a lack of explanation for what data AI algorithms use, or why they may make biased or unsafe decisions. These concerns have given rise to the use of explainable AI, but there’s still a long way before transparent AI systems become common practice.

To make matters worse, AI companies continue to remain tight-lipped about their products. Former employees of OpenAI and Google DeepMind have accused both companies of concealing the potential dangers of their AI tools. This secrecy leaves the general public unaware of possible threats and makes it difficult for lawmakers to take proactive measures ensuring AI is developed responsibly.

AI-powered job automation is a pressing concern as the technology is adopted in industries like marketing, manufacturing and healthcare. By 2030, tasks that account for up to 30 percent of hours currently being worked in the economy could be automated.

Social manipulation also stands as a danger of artificial intelligence. This fear has become a reality as politicians rely on platforms to promote their viewpoints.

A troll army could try to capture the votes of users during an election. Online media and news have become even murkier in light of AI-generated images and videos, AI voice changers as well as deepfakes infiltrating political and social spheres.

These technologies make it easy to create realistic photos, videos, audio clips or replace the image of one figure with another in an existing picture or video.

As a result, bad actors have another avenue for sharing misinformation and war propaganda, creating a nightmare scenario where it can be nearly impossible to distinguish between credible and faulty news.

No one knows what’s real and what’s not - You literally cannot believe your own eyes and ears

In addition to its more existential threat, a prime example of how AI will adversely affect privacy and security is the use of facial recognition technology in offices, schools and other venues.

Besides tracking a person’s movements, governments may be able to gather enough data to monitor a person’s activities, relationships and political views.

Police departments are embracing predictive policing algorithms to anticipate where crimes will occur. The problem is that these algorithms are influenced by arrest rates, which disproportionately impact Black communities.

Police departments then double down on these communities, leading to over-policing and questions over whether self-proclaimed democracies can resist turning AI into an authoritarian weapon.

AI systems often collect personal data to customize user experiences or to help train the AI models themselves (especially if the AI tool is free). Data may not even be considered secure from other users when given to an AI system. One bug incident that occurred with ChatGPT in 2023 “allowed some users to see titles from another active user’s chat history,” according to OpenAI. Similarly, the Meta AI app offers a public “Discover” feed that lets users see other people’s personal chats made with the chatbot. And in 2025, it was found that the Grok chatbot made user conversations searchable on search engines like Google, Bing and DuckDuckGo without warning, with Google estimated to have indexed over 370,000 Grok conversations. As for voice data, AI-powered devices may track and process every voice recording that’s shared with them, including for products like Amazon’s Alexa and various wearable devices. While there are laws present to protect personal information in some cases, there is no explicit law that protects citizens from data privacy harm caused by AI.

Various forms of AI bias are detrimental too. AI bias goes well beyond gender and race. In addition to data and algorithmic bias, AI is developed by humans — and humans are inherently biased. The narrow views of individuals have culminated in an AI industry that leaves out a range of perspectives.

If companies refuse to acknowledge the inherent biases baked into AI algorithms, they may compromise their initiatives through AI-powered recruiting. The idea that AI can measure the traits of a candidate through facial and voice analyses is still tainted by racial biases, reproducing the same discriminatory hiring practices businesses claim to be eliminating. Widening socioeconomic inequality sparked by AI-driven job loss is another cause for concern, revealing the class biases of how AI is applied.

Along with technologists, journalists and political figures, even religious leaders are sounding the alarm on AI’s potential pitfalls. AI’s ability to be misused, and “create statements that at first glance appear plausible but are unfounded or betray biases.” this could bolster campaigns of disinformation, distrust in communications media, interference in elections and more — ultimately increasing the risk of “fueling conflicts and hindering peace.”

As is too often the case, technological advancements have been harnessed for warfare. When it comes to AI, some are keen to do something about it before it’s too late: In a 2016 open letter, over 30,000 individuals, including AI and robotics researchers, pushed back against the investment in AI-fueled autonomous weapons. If any major military power pushes ahead with AI weapon development, a global arms race is virtually inevitable, and the endpoint of this technological trajectory is obvious: Lethal autonomous weapon systems (LAWS), which are designed to locate and destroy targets on their own. Because of the proliferation of potent and complex weapons, some of the world’s most powerful nations have given in to anxieties and contributed to a tech cold war. Many of these new weapons pose major risks to civilians on the ground, but the danger becomes amplified when autonomous weapons fall into the wrong hands. Hackers have mastered various types of cyber attacks, so it’s not hard to imagine a malicious actor infiltrating autonomous weapons and instigating absolute armageddon. If political rivalries and warmongering tendencies are not kept in check, artificial intelligence could end up being applied with the worst intentions. Some fear that, no matter how many powerful figures point out the dangers of artificial intelligence, we’re going to keep pushing the envelope with it if there’s money to be made.

The financial industry has become more receptive to AI technology’s involvement in everyday finance and trading processes. As a result, algorithmic trading could be responsible for our next major financial crisis in the markets. While so-called “AI trading bots”aren’t clouded by human judgment or emotions, they also don’t take into account contexts, the interconnectedness of markets and factors like human trust and fear. These algorithms then make thousands of trades at a blistering pace with the goal of selling a few seconds later for small profits. Selling off thousands of trades could scare investors into doing the same thing, leading to sudden crashes and extreme market volatility. Instances like the 2010 Flash Crash and the Knight Capital Flash Crash serve as reminders of what could happen when trade-happy algorithms go berserk, regardless of whether rapid and massive trading is intentional. This isn’t to say that AI has nothing to offer to the finance world. In fact, AI algorithms can help investors make smarter and more informed decisions on the market. But finance organizations need to make sure they understand their AI algorithms and how those algorithms make decisions. Companies should consider whether AI raises or lowers their confidence before introducing the technology to avoid stoking fears among investors and creating financial chaos.

As AI technology has become more accessible, the number of people using it for criminal activity has risen. Voice cloning has become an issue, with criminals leveraging AI-generated voices to impersonate other people and commit phone scams. Online predators can also now generate images of children, making it difficult for law enforcement to determine actual cases of child abuse. And even in cases where children aren’t physically harmed, the use of children’s faces in AI-generated images presents new challenges for protecting children’s online privacy and digital safety. Chatbots themselves can also negatively impact child safety. In 2025, it was found that Meta AI had policies that allowed the chatbot to have romantic conversations with children. There are also stuffed toys with built-in AI chatbots marketed for young children, where collected data can be sent to third-party companies like OpenAI or Perplexity AI. With AI causing safety concerns for children, this has even led California Attorney General Rob Bonta to warn 12 top AI companies that if they “make choices that lead their technology to harm children, [they] will be held accountable to the fullest extent of the law” in the state of California.

Overinvesting in a specific material or sector can put economies in a precarious position. Like steel, AI could run the risk of drawing so much attention and financial resources that governments fail to develop other technologies and industries. Plus, overproducing AI technology could result in dumping the excess materials, which could potentially fall into the hands of hackers and other malicious actors.

There also comes a worry that AI will progress in intelligence so rapidly that it will become conscious or sentient, and act beyond humans’ control — possibly in a malicious manner. Alleged reports of this sentience have already been occurring, with one popular account being from a former Google engineer who stated the AI chatbot LaMDA was sentient and speaking to him just as a person would. Anthropic research also found that some leading AI models are willing to resort to malicious behaviors to reach their goals, including blackmailing officials and leaking sensitive information when faced with replacement or conflicting assigned goals in a stress-testing simulation. Google itself also recognizes the risk of AI systems acting against human commands, as the company expanded its Frontier Safety Framework to address potential future scenarios where AI models may interfere with operators’ abilities to direct, modify or shut down their operations. As AI’s next big milestones involve making systems with artificial general intelligence, and eventually artificial superintelligence, calls to completely stop these developments continue to rise.

Generative AI has ignited a fierce legal and ethical battle over the ownership of creative works. Because large language models and image generators are trained on massive datasets scraped from the internet, they often ingest copyrighted books, articles, code and artwork without the original creators’ consent or compensation. This has led to several high-profile lawsuits from authors, visual artists and news organizations who argue that AI companies are regurgitating their intellectual property to create competing content. AI intellectual property infringement harms artists by undermining their livelihood, as they are forced to compete against machines trained on their own work and can produce new material at much faster speeds. For AI companies, IP infringement creates massive legal vulnerabilities that are still being worked out in the courts. Eventually, it can also lead to model collapse by disincentivizing human creators and forcing models to train on their own repetitive outputs.