San Francisco, Sept 19 : Artificial intelligence company Anthropic and consulting and technology services firm Accenture have announced a partnership aimed at expanding independent evaluation of advanced AI models, as scrutiny around the safety and reliability of frontier systems increases.
Under the partnership, the two companies plan to commit at least $1 billion each over the next five years, taking the combined planned investment to a minimum of $2 billion.
The initiative is expected to strengthen the infrastructure and expertise needed to assess increasingly capable AI systems before and after their deployment. The companies said the programme will focus on independent evaluation, with the aim of improving understanding of how advanced models perform and identifying potential safety and reliability risks.
Anthropic has been among the AI developers placing significant emphasis on model safety and evaluation as increasingly capable systems are adopted across businesses and other sectors.
The partnership comes at a time when AI developers are facing increasing expectations from regulators, businesses and researchers to demonstrate that advanced models operate reliably and can be deployed responsibly.
As AI systems become more capable and are integrated into critical business processes, independent testing and evaluation are gaining importance. Such assessments can examine model performance, potential vulnerabilities and the effectiveness of safeguards.
Accenture’s involvement is expected to add large-scale technology, consulting and implementation capabilities to the evaluation effort, while Anthropic will bring its experience in developing and testing frontier AI models.
The companies said the investment will help build capacity for AI evaluation over the next five years, reflecting the growing demand for independent assessment as the technology develops.
The announcement comes amid a broader industry push to establish stronger testing and safety practices for advanced AI models, with companies, governments and researchers exploring ways to improve transparency, reliability and accountability in the sector.
