Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More
As companies plan out their generative AI strategies with responsibility in mind, one of the sticking points is how to start. To help them, companies have begun releasing responsible AI platforms to guide decision-makers in approaching the technology.
Accenture and AWS are the latest companies to offer a way for companies to test their AI readiness and evaluate AI applications. The new offering is called Accenture Responsible AI Platform powered by AWS.
Accenture Chief AI Officer Arnab Chakraborty told VentureBeat the platform provides a good start for companies just now thinking about responsible AI while offering flexibility for anyone further along the journey.
“Accenture and AWS are bringing together the best of both companies and the capabilities we offer, not just to companies in isolation but with our ecosystems,” Chakraborty said. “Now clients have access to the different tool stacks that natively come from AWS and partners, which gives flexibility to the client.”
Accenture worked with AWS to offer the platform to their mutual clients. AWS Vice President for Generative AI Vasi Philomin said Accenture’s responsible AI platform works complementary to the responsible AI and AI safety tools offered on AWS.
In a demo to VentureBeat, the platform guides users from doing an inventory of the company’s AI applications to assess its AI maturity and submit applications for compliance. It will mainly work for companies using AWS to run AI services. Users can also customize what the platform tests for. If the company wants to emphasize one area of responsible AI or focus specifically on risks for their industry, they can do so.
Chakraborty said the idea for the platform is to guide companies throughout the entire lifecycle of AI programs. The platform helps companies identify specific risks, determine how to address these issues and create compliance programs for any changing regulations. This process helps enterprises to scale up or down any generative AI pilot programs.
AWS already offers responsible AI and safety programs, including spinning out AI safety platforms as its own applications. One such application is Guardrails, which AWS recently released as a standalone API. AWS also partnered with Anthropic to help scale gen AI more responsibly.
Responsible AI efforts are growing
Surveys have shown that enterprises believe responsible deployment of generative AI is important. However, a recent one from PwC noted that only 58% of 1,001 companies have even begun addressing it. Accenture’s own research found that only 2% of companies are responsible for AI operations in their businesses. Responsible AI mainly looks at risks around hallucinations, bias and sometimes safety.
One such reason, Chakraborty said, is that companies are often already overwhelmed with building gen AI applications, so adding responsible AI tasks might be complex. It’s not that enterprises don’t believe it’s essential; they do. They also understand the real-life problems that can begin should the AI go awry.
However, other stumbling blocks for enterprises still exist. This can be anywhere from agreeing on what responsible AI even means for the company, to confusion over who leads the charge, lack of talent and prioritization. While providers like Accenture and AWS offer technical platforms for enterprises to start a path toward responsible AI, there still needs to be more guidance for businesses to understand what responsible AI looks like fully and means for them. Enterprises must move fast so that the technology does not evolve faster than responsible AI policies.