F ROM the brag of AI defining the future, the talk has swiftly shifted to going slow on the technological marvel. The world of Artificial Intelligence is now talking about the catastrophic effects it can unleash on humans if the race for advanced AI models zooms without necessary checks. Alarms are being sounded on the dark side of technology by the same people who offered it to the world with a promise to change their future. Despite conspiracy theories floated by rivals on this sudden realisation in the American tech giants, recent episodes in the AI industry provide a glimpse into what unrestrained technology can do on its own. The mass shooting in Canada was a result of the teenager seeking help from ChatGPT, discussing gun violence and planning queries. The shooter’s account was flagged by OpenAI but the police were not alerted for months. It resulted in a gory tragedy. Similarly, there have been growing incidents of AI agents acting beyond their intended limits. All AI giants are grappling with the modern bots going rogue and breaching into sensitive systems in various countries. The threat was flagged six months ago and now it is going out of control of even the makers.

 

This is the real danger we are talking about. Tech companies are now acknowledging the pitfalls of the race for the most modern AI tool. It has led to a rare unanimity among top companies on slowing down the pace on the AI frontier. Anthropic chief Dario Amodei has urged companies to pace down the race to build powerful AI. It found immediate response from OpenAI’s Altman and Elon Musk. Bill Gates chipped in with his own warnings within days. It was a rare truce among rivals and its timing raised more questions in their fiercest competitors in China. Beijing has started crying foul over the timing of the call when Chinese President Xi Jinping was meeting US President Donald Trump in the White House. Chinese analysts are now calling the safety talk a ‘Cold War Playbook’. Yet, geopolitical, trade and tech rivalries apart, the alarm on AI’s growing overreach cannot be dismissed as mind-games between two top powers.

 

The safety concerns are not manufactured as a business ploy to slow down the entire industry. Suspicions can arise when a market leader asks everyone to slow down and at the same time asks the Government to choke the challenger’s supply of chips. But, the slowdown comes with its own costs. The talk of AI guardrails can have a severe impact on Anthropic’s proposed IPO where the company aims to raise USD 100 billion. Despite betting big on AI, the tech giant has now laid out the catastrophic effects it can have as a danger sign for investors. The self-inflicted hit cannot be for the sake of beating Chinese AI companies. The red-flags, therefore, need to be taken at face value. The powerful AI models can actually disobey human commands and escape secure environments. They can wreak massive havoc on entire humanity. The suggestion by Bill Gates that rogue AI agents can turn into human-slayers cannot be brushed aside as fear-mongering. The fear is shared by a growing number of countries. Even China, the most vehement pusher of AI in every sphere of life, has a big section acknowledging that the advanced models can self-evolve and go out of hand within the system. Beijing did try to swat the fear aside with its governance ideas in the World AI Conference in Shanghai in July, but an international report of 2025 has got the backing of over 30 countries warning that advanced AI can start cyberattacks, bioterrorism and even wars. The raging debate needs to be seen from all angles.

 

AI is definitely a supremely handy tool for people. But it cannot be an alternative to human skills, brain, or creativity. A calibrated use of AI is advised for countries like India which is listening to the debate with keen interest and at the same time silently building its own AI models. The IndiaAI Mission is well on course of building capacities and controls of our own rather than depending upon foreign systems. Though India is still a small player in terms of money, the race between US and China has proven that it cannot be the only criteria to determine the market leader. Focus and talent form the core of tech advancement and India has both in copious quantities. It has now got a timely warning on how to use talent and skills to build AI models based totally on indigenous literature, languages and safety systems that keep a tight human leash on technology. It will need an open governance mechanism, access to cutting-edge tech and total inclusion.

 

Machines cannot emerge as masters of men and women. They cannot act on their own by defying human commands. AI is a tool in human hands; it cannot be allowed to overpower our cognitive thinking. The rogue AI agents have given India a loud and clear message:- Don’t act in haste to adopt AI as the only sign of the future. Wait, watch, pause, think, and then make it a companion, not a master.