OpenAI, renowned for its collaboration with enterprises such as Klarna, Morgan Stanley, Oscar, Salesforce, and Wix, is furthering its commitment to assisting businesses in constructing AI solutions from the ground up and ensuring safe deployment across their organizations and products. The company is expanding its support for enterprises with a suite of new features tailored for both large corporations and developers experiencing rapid growth on its platform.
Enhanced Enterprise-Grade Security
Among the latest enhancements is the introduction of Private Link, offering customers a method to establish direct communication between Azure and OpenAI while minimizing exposure to the open internet. Complementing this is the rollout of native Multi-Factor Authentication (MFA), designed to aid compliance with increasingly stringent access control requirements. These additions join OpenAI’s existing array of enterprise security features, including SOC 2 Type II certification, single sign-on (SSO), data encryption at rest using AES-256 and in transit using TLS 1.2, and role-based access controls. Furthermore, the company offers Business Associate Agreements catering to healthcare companies requiring HIPAA compliance and maintains a zero data retention policy for API customers with qualifying use cases.
Enhanced Administrative Control
OpenAI introduces the Projects feature, affording organizations greater granularity and oversight over individual projects within its platform. This includes the ability to tailor roles and API keys to specific projects, regulate the availability of models, and impose usage- and rate-based limits to manage access and prevent unexpected overages. Project owners gain the capability to create service account API keys, providing project access without being linked to individual users.
Assistants API Improvements
Several updates have been implemented to the Assistants API, enhancing retrieval accuracy, flexibility in model behavior, and cost control. Key features include improved retrieval with ‘file_search,’ supporting up to 10,000 files per assistant, streaming support for real-time conversational responses, and the introduction of ‘vector_store’ objects for simplified file management and billing. Additionally, support for fine-tuned GPT-3.5 Turbo models has been integrated into the API.
More Options for Cost Management
OpenAI introduces discounted usage on committed throughput for customers with sustained levels of tokens per minute (TPM) usage on GPT-4 or GPT-4 Turbo. Furthermore, the company unveils the Batch API for running non-urgent workloads asynchronously, offering reduced pricing and higher rate limits.
Looking Ahead
OpenAI pledges to continue enhancing its offerings with a focus on enterprise-grade security, administrative controls, and cost management. Interested parties can access further information on these launches via the company’s API documentation or engage with its team to explore custom solutions tailored to their enterprise needs.
