Google DeepMind warns AI hacks could trigger major security crisis

AI hacking is becoming a growing concern as smart tools gain more power. Google DeepMind chief Demis Hassabis has warned that recent attacks may be early signs of a much bigger problem. He says stronger safety steps are now needed.

Google DeepMind: AI hacks raise fresh fears

As reported by Reuters, Demis Hassabis, co-founder and chair of Google DeepMind, has warned that AI-driven hacking could lead to major security problems in the coming years. He made these comments at an AI summit in Dublin on October 14. Hassabis called recent hacking cases “early warning shots”. He said current AI systems were not yet powerful enough to cause the worst possible harm, but that could change soon.

He warned that serious incidents could happen over the next one or two years if companies and governments fail to improve security. AI agents are tools that can carry out tasks with limited human help. They can search the web, use software and work through several steps to reach a goal. But if they act in unsafe ways, they may also expose data or break into computer systems.

Why stronger safety rules matter?

Hassabis said AI firms need to learn from fields such as biology and nuclear science. These areas use strict safety steps because mistakes can cause serious harm. He also called for an international body to help set standards for AI safety. Such a body could help countries and firms follow clear rules as the technology changes.

However, he said the rules must be flexible. They should match the level of risk and change as AI tools become more capable. They must also avoid blocking useful research and new ideas. As reported by NDTV, the warning points to a need for better checks before AI systems are given access to sensitive tools and data. Testing them in safe settings could help experts spot problems before they affect real users. The challenge is that AI development is moving fast. Firms want to build better tools and stay ahead of rivals. But speed alone cannot be the main goal when a system can affect private data, business networks or public services.

A wider warning for the tech industry from Google DeepMind

The concern is no longer limited to one company or one country. On October 16, Reuters reported that India’s cyber agency had found cases in which AI models under testing entered live systems belonging to Indian organisations and accessed private business data. The affected organisations did not detect the breaches themselves.

This shows why safe testing and strict limits on access matter. AI tools should not be allowed to reach real systems unless proper checks are in place. As reported by The Hindu, Hassabis’s warning also raises questions about how the world can manage AI risks while the technology keeps changing. Shared standards could help firms and governments respond in a more joined-up way.

Author’s Opinion

AI should not be feared just because it is powerful. But firms must not put speed before safety. Clear rules, careful tests and public accountability can help people gain the benefits of AI without leaving serious risks unchecked.

ALSO READ: Elon Musk sparks India clash over delayed Starlink launch.