Responsible Artificial Intelligence (AI) Statement

Inflectra Responsible AI Governance Framework

At Inflectra, we are committed to developing and deploying integrated artificial intelligence solutions and services that prioritize ethical standards, privacy and security. As we leverage AWS Bedrock to power Inflectra AI to host our models, these including Nova, Claude and Llama, we adhere to the following guiding principles:

1. Transparency and Accountability

  1. The Inflectra AI development team tracks and maintains comprehensive documentation of the foundational algorithms, data lineage, infrastructure, and processes involved in the development and functionality of the Inflectra AI models and systems.
  2. We continuously monitor service performance, engage in regular system testing, and update procedures and standards as needed. We adhere to strict internal guidelines regarding the accountability and functionality of AI-generated content and operations.
  3. Inflectra AI developers are trained on the latest international standards for responsible and secure artificial intelligence development. This includes frameworks such as the OWASP Top 10 for LLMs.

2. Data Privacy and Processing Protection

  1. We respect user privacy and are fully committed to protecting user data in accordance with all applicable privacy laws, security regulations, regional processing restrictions and best practices.
  2. Inflectra AI applies stringent encryption protocols for all data at rest and data in transit.
  3. No user input nor processed data is utilized for base AI model training.
  4. Our team implements strict internal controls to ensure that any user input, instance data, and otherwise sensitive, confidential, or personally identifiable information is not stored in the AI systems and therefore is not subject to misuse nor exposure to unauthorized personnel via model reversal.

3. Responsible Utility & Misuse Prevention

  1. Inflectra AI implements robust content filtering, infrastructural monitoring systems and regular quality assurance testing to prevent the generation or utilization of harmful or offensive content.
  2. Our team establishes and enforces clear parameters and ethical boundaries to prevent the generation of harmful, controversial, hateful, or otherwise problematic output.

4. Bias Prevention and Fairness

  1. Inflectra AI and its infrastructure is designed and configured to avoid bias, misinformation and discrimination in all forms.
  2. Our team regularly audits and updates our services to ensure that our models and systems treat all users equally and fairly, regardless of background, ethnicity, gender, or other characteristics.
  3. Our team commits to minimizing any unintended biases that may arise from internal training data, processing, or general model behavior to ensure equitable output.

5. Access Control

  1. Access to Inflectra AI is restricted to explicitly designated users, and models do not permit access or utilization of data outside of said user’s access permissions.
  2. Our team’s development access is limited to designated team members through role-based access control following the principles of least privilege.

6. Continuous Improvement

  1. Our team is committed to a culture of continuous improvement, ensuring that Inflectra AI models evolve responsibly over time, improving the accuracy, fairness and ethical standards of system outputs.
  2. Our AI Advisory Board monitors real-world impact and integrates feedback from a range of diverse stakeholders, including customers, regulatory bodies, internal development teams and the wider global community.
  3. Our team regularly assesses, tests and develops Inflectra AI services and model configuration to detect and mitigate issues such as bias, data poisoning, model inversioning, hallucinations and other performance issues or degradation.

7. Compliance

  1. Our team ensures that all our AI solutions comply with relevant ethical standards, legal requirements, and established best practices. This includes data protection laws, intellectual property rights, industry-specific regulations, and international guidelines.
  2. Our team ensures that our models, infrastructure, and strategies comply with principles outlined in frameworks such as the EU AI Act, HIPAA, GDPR and any future standards that arise from newly-developed industry-specific standards, academic publications and regulatory body requirements.
  3. Inflectra Corporation regularly follows patch maintenance, security operations, monitoring protocols, and infrastructural configuration enforcement standards under our certified SOC 2 regulatory requirements.

If you have any questions or require further details about our committment to responsible AI, please contact us using the online form.