Breaking

Anthropic and Accenture Form Embedded AI Safety Evaluation Partnership

September 19, 2026 at 11:03 EDT

Foundation Models
AI Agents
Security

Anthropic and Accenture said they will create an embedded evaluation team inside Anthropic, giving outside specialists employee-like access to assess frontier AI safety while Claude models continue to be developed and released. The companies said each expects to invest at least $1 billion over five years in AI safety capacity, with Accenture's Faculty unit leading work on model red-teaming, alignment assessments and safeguard testing.

September 18, 2026 · Anthropic and Accenture

Embedded evaluators move inside Claude development

The companies plan a new safety evaluation team with employee-like access inside Anthropic, backed by at least $1 billion from each company over five years.

$1B+

Expected from Anthropic for AI safety capacity

$1B+

Expected from Accenture over the same period

5 years

Capacity-building window for the safety push

The funding scale is matched

$1B+

Anthropic

$1B+

Accenture

Equal-height stacks show each company expects to commit at least the same dollar amount.

What the embedded team is meant to do

Evaluate modelsInspect frontier systems as they develop.

→

Red-team risksProbe safeguards, security and behavior.

→

Report blind spotsAssess practices and identify gaps.

Why supporters see value

Employee-like access could let reviewers see training, deployment and safeguard decisions earlier than conventional outside testing.

What remains unsettled

Standards for access, reporting, independence and long-term funding have not yet been established.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.