SAN FRANCISCO — OpenAI is committing $1 billion in subsidized access to advanced artificial intelligence cybersecurity tools, training and technical assistance in an effort aimed partly at helping smaller financial institutions and other resource-constrained organizations defend themselves against increasingly sophisticated cyberattacks.
In a blog post, the company said its new “Daybreak for Frontline Defenders” initiative will initially prioritize organizations in the United States responsible for essential services, including water and electric utilities, state and local governments and community and regional banks.
Notably for credit unions, however, OpenAI’s announcement specifically identifies community and regional banks but does not mention credit unions or indicate whether they are included among the financial institutions eligible for the subsidized access. The CU Daily has contacted OpenAI seeking comment.
OpenAI also said it has been meeting with frontline defenders that include community banks, but again made no reference to credit unions. Its Daybreak website similarly lists community banks among organizations that can register their interest in receiving support.
$1B Targeted for Use Over Six Months
OpenAI said it is targeting the $1 billion commitment to be consumed over the next six months, beginning in the United States and later expanding to partner countries.
The initiative includes:
- Subsidized access to OpenAI’s Daybreak cyber models and products.
- Cybersecurity training and technical support.
- Partnerships intended to help organizations deploy AI-based defensive capabilities.
- More than 35 enterprise products and partner-operated services available through the Daybreak Defense Network.
- A new pilot with the Multi-State Information Sharing and Analysis Center, or MS-ISAC, initially focused on public-sector and water-system cybersecurity.
OpenAI said the initiative is aimed particularly at organizations that carry significant cybersecurity responsibilities but lack the budgets, staffing and specialized expertise available to large enterprises.
AI Could Help Find, Fix Vulnerabilities
OpenAI said Daybreak can be used by cybersecurity teams to review legacy computer code, analyze suspicious activity, identify and validate vulnerabilities, prioritize risks and develop and test fixes.
The company launched Daybreak earlier this year to provide verified public- and private-sector cybersecurity professionals with advanced AI capabilities for authorized defensive activities.
Daybreak Blue uses OpenAI’s primary models for common defensive cybersecurity work, while Daybreak Red provides approved organizations with access to specialized cyber models for more sensitive and technically demanding tasks.
According to OpenAI, thousands of cybersecurity professionals across 2,000 approved organizations and workspaces are already using Daybreak, including cybersecurity companies, defense organizations and law enforcement agencies.
The company’s broader Daybreak platform is designed to help defenders create a continuous process of identifying, validating, assigning, remediating and documenting vulnerabilities while maintaining human oversight.
OpenAI Warns AI Will Also Strengthen Attackers
The initiative comes as OpenAI is warning that the same advances in AI that can strengthen cyber defenses are expected to make attacks more widespread and sophisticated.
The company described the current period as a “defender’s window” in which organizations have an opportunity to use advanced AI to identify and close security weaknesses before attackers can take advantage of increasingly capable models.
OpenAI said the effort follows a call for collective cybersecurity action supported by more than 150 organizations across the technology, cybersecurity, critical infrastructure, finance and AI sectors.
The company said it wants to concentrate resources on organizations that often must protect complex and aging systems despite having fewer cybersecurity resources than large corporations.
Pilot Will Pair AI With Training
As part of the initiative, OpenAI is partnering with MS-ISAC on a pilot involving state, local, tribal and territorial governments and water-system defenders.
Participants will receive Daybreak access along with training and hands-on assistance designed to help them validate vulnerabilities, determine which problems pose the greatest risks and coordinate remediation.
OpenAI said the pilot is intended to create a model that eventually could be expanded to other organizations.
Separately, the Daybreak Defense Network is rolling out more than 35 products and partner-operated services that incorporate OpenAI’s cyber models into cybersecurity tools and workflows already used by enterprises.
Previous Assistance Followed Water-System Attacks
OpenAI said the $1 billion commitment builds on assistance it previously provided following cyberattacks against U.S. water systems.
The company said affected states and utilities were offered up to $1 million in no-cost API credits, Daybreak access and technical assistance. Those organizations used the resources to review computer code and system configurations, validate findings, develop patches and confirm fixes while keeping water systems operating, according to OpenAI.
OpenAI said its ultimate objective is not simply to use AI to discover more cybersecurity vulnerabilities, but to shorten the time required to validate and repair them.
