evidoria

← Back to browse

Good practice Imported

Daybreak for America — OpenAI and MS-ISAC Launch a Subsidised AI Cyber-Defence Pilot for U.S. State and Local Governments

United States of America · East Greenbush · See the United States of America profile

Evidence: Descriptive / self-reported Top 62% 53/100 · Ask Evidence Copilot about this practice

OpenAI pledged $1 billion in subsidised access to its Daybreak cybersecurity tools and launched an MS-ISAC pilot pairing training with AI threat-hunting for under-resourced U.S. state, local and water-utility defenders after a wave of attacks in mid-2026.

1 USD billion
Global subsidised-access commitment
1 USD million
No-cost credits offered to states/utilities
100+ incidents
Water/wastewater cybersecurity incidents reported (July 2026)
2,000 organisations
Organisations/workspaces already using Daybreak

Details

Maturity
Pilot
Promoter
OpenAI and the Multi-State Information Sharing and Analysis Center (MS-ISAC, Center for Internet Security)
Period
2026
Keywords
cybersecurity, critical infrastructure, water utilities, public-sector IT

Context

On 3 September 2026, OpenAI announced Daybreak for Frontline Defenders, a $1 billion global commitment of subsidised access to its Daybreak cybersecurity models, training and technical support for resource-constrained cyber defenders.

Objectives

As part of the commitment, OpenAI launched a pilot with the Multi-State Information Sharing and Analysis Center (MS-ISAC), run by the nonprofit Center for Internet Security, to help under-resourced public-sector and water-system defenders validate and prioritise security findings and coordinate remediation.

Activities

The subsidy is meant to be drawn down over roughly six months, and OpenAI separately offered $1 million in no-cost credits and support to states and utilities after water and wastewater systems in seven states reported more than 100 cybersecurity incidents involving programmable logic controllers in July 2026.

Results

OpenAI states that 2,000 approved organisations and workspaces, including law-enforcement and defense entities, already use Daybreak models.

Conclusions

The pilot was announced only two weeks before this review, so no independent results, adoption numbers within the MS-ISAC cohort, or measured security outcomes are yet available; it is documented here on the strength of its funding commitment and named scope, not demonstrated impact.

Implementation
Indicative cost
Very high (> €5M)
Time to results
Short (< 1 year)
Staffing & skills
OpenAI, Multi-State Information Sharing and Analysis Center (MS-ISAC), run by the nonprofit Center for Internet Security, state, local and water-utility defenders in the initial cohort

Conditions for success

  • an existing MS-ISAC base of thousands of public-sector members and 2,000 organisations already using Daybreak models
  • subsidised access paired with guided training rather than tooling alone

Common failure modes

  • announced only weeks before this review, with no independent results, adoption numbers, or measured security outcomes yet available

Commonly funded by

National / regional programmes Philanthropic / foundation funding

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Similar practices you may find useful