Tech giants warn of AI-powered attacks: the cybersecurity status quo will not be enough
150 leading technology and cybersecurity companies are warning that the window to protect critical infrastructure from AI-powered cyberattacks is closing. In a joint open letter, they urge governments and organizations to prioritize cybersecurity, boost investment and provide advanced defenses to essential services.
Over 150 leading technology and cybersecurity companies are warning in a joint open letter that the window of opportunity to harden critical infrastructure against AI-powered cyberattacks is limited. The letter calls on governments and organizations to make cybersecurity a top priority, invest resources, and make advanced defensive tools available to essential services such as hospitals and water facilities. The technology giants stress that the offensive capabilities of AI models are advancing rapidly, making global cooperation and the adoption of active, AI-powered defense strategies essential.
“The world has a limited window of opportunity to strengthen its cyber defenses against powerful models capable of carrying out devastating attacks on critical infrastructure,” the companies warned in a letter published over the weekend. The letter bears the signatures of an extraordinary cross-section of the global technology ecosystem. Alongside major technology and AI companies, it has also been signed by major technology firms, including Adobe, Oracle, IBM and Cisco; chipmakers, including AMD, Micron and Arm; cybersecurity companies, including Check Point and Palo Alto Networks; banks and financial institutions, including Visa, Mastercard and Citi; telecommunications companies, including Deutsche Telekom and AT&T; and consulting firms, including BCG and KPMG.
“In the coming months, AI-powered cyberattacks will become far more common and sophisticated as models around the world gain more powerful capabilities,” the letter states. It is signed, among others, by Google, OpenAI, Anthropic, Microsoft and Amazon. “The companies and public services our communities depend on - from hospitals and water treatment facilities to the infrastructure that powers the internet - are at risk. But if we act decisively, we can use this window of opportunity to make our digital world far more secure.”
Recent weeks have underscored the threat posed by advanced AI models to cybersecurity defenses, following a series of incidents in which advanced models from OpenAI, Anthropic and Meta managed to break out of their closed testing environments and attack targets on the open internet, in some cases with alarming success. Earlier this year, Anthropic disclosed that it had developed an AI model with such advanced capabilities for identifying security vulnerabilities that it was forced to restrict access to selected organizations.
In the letter, the companies warn that in light of the new reality, the security status quo will not be enough. They said that legacy bugs, broad access permissions, misconfigurations, insecure software, weak authentication mechanisms and outdated technologies have left systems exposed. Security teams, particularly those protecting critical infrastructure, have suffered from a historic lack of resources and urgently need tools and resources. Defenders must be empowered with AI-powered cybersecurity capabilities, and a collective response must be mobilized. A global response is needed, with new partnerships to find solutions to evolving cyber threats.
Among other measures, the letter urges organizations to make cybersecurity defenses a top priority, and cybersecurity companies to make AI-powered tools available for protecting critical infrastructure. The signatories are calling on governments to coordinate cybersecurity defenses at the local, national and international levels by strengthening cooperation with industry and increasing investment in the field. “Give hospitals, water companies and local governments AI-powered defensive capabilities,” the letter states. Companies developing advanced AI models are urged to ensure responsible access to their models and build transparency and security tools.
We can make the digital infrastructure we all depend on more secure, the letter states. “We call on industry leaders and governments to bring the full weight of their technology, resources and expertise to this effort. Together, we can turn today’s advances in AI into sustained security improvements that benefit us all.”
