Accenture and Anthropic have reached an agreement to collaborate on the creation of a team of integrated evaluators who will work alongside Anthropic’s internal teams and security partners. Its objective will be to evaluate and subject the AI models to different tests to identify possible vulnerabilities, perform alignment evaluations and verify the effectiveness of their safeguards.
As part of this collaboration, both Accenture and Anthropic plan to invest at least $1 billion each over the next five years in initiatives related to artificial intelligence security. The agreement seeks to strengthen the evaluation mechanisms of the models as they acquire an increasingly relevant role in companies and organizations.
Faculty will complement Accenture’s capabilities
Accenture will contribute to this initiative its decades of experience and investment in responsible AI, in addition to the technical and security capacity of Faculty, a company specialized in applied artificial intelligence that was acquired by the consulting firm. Faculty’s experience will complement Accenture’s capabilities and collaborate with Anthropic in the development of new evaluation methodologies.
Faculty has extensive experience in evaluating and testing AI models for some of the world’s leading artificial intelligence laboratories. The company has also participated in the development of complex systems designed from the beginning under security and responsibility criteria.
His experience extends to sectors such as public administration, defence, health and infrastructure. Among its projects is the development of the Early Warning System of the British National Health Service (NHS) during the COVID-19 pandemic. This experience will now be applied to joint work with Anthropic to evaluate the behavior and safety of AI models.
“Accenture is bringing together a dedicated team with extensive experience in AI, security and cross-sectors to work alongside Anthropic,” said Julie Sweet, president and CEO of Accenture. “Security requires both deep technical knowledge and a clear understanding of how AI is used in the real world.”
According to Sweet, integrated evaluation is an emerging area within artificial intelligence security. “We look forward to collaborating with Anthropic to help accelerate the development of integrated evaluators, which we believe will be an important part of the security ecosystem in the future,” he said.
Integrated Cybersecurity Evaluators
For his part, Dr. Marc Warner, Chief Technology Officer of Accenture and CEO of Faculty, has highlighted that security must be part of the development of AI systems from their early stages. “Faculty, which is now part of Accenture, was born with the conviction that AI must be safe by design, and not safe by chance,” he explained.
Both companies plan to invest at least $1 billion each.
Warner has noted that this principle has guided Faculty’s work with governments, defense and health systems around the world. In this context, he considers that the collaboration with Anthropic fits with the company’s trajectory and will allow it to expand its capacity to act.
“Joining forces with Anthropic as integrated evaluators is exactly the type of work that Faculty was created to do and, now as part of Accenture, we have the platform necessary to carry it out at a scale capable of generating real impact in the industry,” he concluded.
