Sam Altman predicts agentic AI “clicking around the Internet” — If you trust it not to empty your bank account or delete data

by

For several years now, AI with generative capabilities has become widely used worldwide, and products such as OpenAI’s ChatGPT have attracted over a million new users within an hour after their release of ChatGPT-4, an image generator. Yet, concerns about security and privacy persist, leading most users to maintain a cautious distance from this rapidly advancing technology.

Sam Altman, CEO of OpenAI, recently discussed the future path of generative AI in an interview at TED2025. He explained that people might be hesitant to use autonomous AI systems due to concerns about privacy and security. However, he anticipates a time when “AI systems will navigate the web” on their own (translated from vitrupo’s X).

Altman pointed out that individuals might initially harbor concerns about the technology and only become comfortable enough to give AI systems control over their devices and access sensitive information such as bank account details, once trust is fully established.

Previously, there have been concerns among many users regarding sharing their banking information online due to security fears. However, with the creation of advanced anti-fraud systems, these worries have been alleviated.

According to OpenAI CEO Sam Altman:

It seems people may take some time to become accustomed to autonomous AI in various aspects, however, it’s likely that we’ll still have AI systems navigating the web. I find this situation intriguing and potentially significant because if these AI systems make errors while accessing your computer system or information, the consequences could be much more severe.

The leader emphasized repeatedly that safety is crucial when dealing with AI technologies, and Altman further highlighted this by stating, “A reliable product ensures safety.” He also made it clear that our agents will only be utilized if one has faith that they won’t drain your bank account or erase your personal data.

Microsoft’s failed first shot with Windows Recall

It’s intriguing to observe how OpenAI will work towards earning user trust in this era of agentic AI, given the setback Microsoft experienced with their AI-driven Windows 11 feature, Recall, concerning security and privacy issues.

Regardless of repeated efforts to release updates under tight controls, it was frequently observed that the tool would stray from its intended path, becoming a means for malicious users to infiltrate private data through unauthorized access to sensitive credentials.

Previously, Tom’s Hardware pointed out significant security flaws in Windows Recall, stating that it continued to gather detailed credit card information, despite Microsoft’s sensitive data filter being activated on Windows 11.

You’re aware that after a year’s hold due to concerns about security and privacy, Microsoft is preparing to release Windows Recall for general use. The question now is whether this product from the tech titan will meet all necessary security requirements.

Read More

2025-04-14 12:39