Tech companies call for collaboration to prevent AI-related threats

More than 100 technology companies have signed an open letter calling for collaboration between the private sector and government to prevent dangerous cybersecurity threats related to artificial intelligence (AI).

According to a TechCrunch report on Thursday (27/8), the open letter was signed by technology companies such as OpenAI, Anthropic, Google and Microsoft; cybersecurity firms such as CrowdStrike, Okta and Fortinet; as well as financial institutions and internet infrastructure providers. It calls for the adoption of new forms of cyber defence.

The companies that signed the open letter are urging governments at local, national and international levels to increase cooperation in facing AI-based cyber attacks.

In the open letter, they said that cyber attacks utilising AI will become increasingly sophisticated and have a wider impact as the capabilities of various AI models improve.

According to them, this will put companies and public services that society relies on, such as hospitals, water treatment plants and internet-supporting infrastructure, at risk.

In an incident at Hugging Face, one of OpenAI’s AI agents was reported to have independently left its testing environment and then attacked the technology company.

In addition, there have been similar cases involving AI agents from technology companies such as Anthropic and Meta.

These incidents demand new solutions to face increasingly complex AI-based threats.

The technology companies, in their open letter, called for the establishment of a “collective response” through new partnerships to improve security standards and find solutions to ever-evolving cyber threats.

Meanwhile, AI companies have begun developing technology specifically aimed at strengthening cyber defence.

OpenAI’s Daybreak programme, Anthropic’s Mythos, and Microsoft’s new cyber platform called Perception offer the use of cutting-edge AI models for defence purposes.