AI Policy

Since the debut of ChatGPT late in 2022, our world has changed significantly. AI’s introduction into the public arena turned this technological umbrella term into a global phenomenon and near-universal corporate buzzword. Everyone is getting in on the act, and not without reason. The technologies that fall under that umbrella are proliferating in every facet of our lives, from search and social media to education and medicine. It is no longer a matter of if we will have AI or when we will have AI, but how we will use AI.

In every industry, there are bad actors and false advertisers, but for the rest of us we must decide if we will be thoughtless users of AI or thoughtful users of AI. In light of this, and in hopes of fostering a productive atmosphere in the sensitive security industry in which we work, Garrett has developed the AI policy below.

Garrett Artificial Intelligence Use Policy

  1. Purpose and Scope:
    1. To outline the ways Garrett uses AI in relation to the design and creation of products and services and the ways we ensure the safety and satisfaction of our customers.
  2. Core Principles:
    1. Use AI tools to enhance and extend our design capabilities to build more power and utility into our products.
    2. Avoid unintended consequences by keeping appropriate human intervention and analysis “in the loop”.
    3. Adequately test all AI work products.
    4. Protect sensitive information and customer data.
    5. Above all, to be trustworthy in our products and processes, deserving of the confidence of our end users in Security and Sport applications and the millions of people scanned and protected by Garrett machines each year.
  3. Data Privacy:
    1. Garrett does not expose customer data or sensitive security information to off-prem AI platforms. We will ensure that the algorithms that are used in our products and the product capabilities which relate specifically to detection for security products are never exposed to “bad actors” or to the public.
    2. As with all our product software, we will be diligent in encrypting our code to offer products that enable our most advanced proprietary technology in all use cases.
  4. Human Oversight:
    1. The results of AI-enhanced design exercises are never used in products through automated or unsupervised means and methods. All AI outputs that are used in product or services design activities are vetted by skilled, experienced human agents. Garrett does extensive product testing in the lab and in the field to ensure that any AI work product is effective and suitable for use in our products.
  5. Risk Management:
    1. Garrett mitigates the risk of using AI in development activities by:
      1. Using on-prem development resources where needed for privacy reasons.
      2. Performing extensive testing on AI generated IP and techniques.
      3. Applying appropriate industry standards to our products for the purpose of assessing health and safety, detection, security, and operational effectiveness.
      4. Using AI in the development of security products in such a way that the product capability increases but is never left “open loop”. Product software is always left in a “locked down” state in our products so that product performance is always based on initial development activities, understood by the end user through their use of application testing, and verifiable.
      5. To protect data that may be interpreted by AI-built algorithms, all our cameras used in Garrett products are NDAA compliant.
  6. Responsible Use:
    1. Garrett will use AI in the design phase of our security products, then “lock down” our software so that it performs at a consistent level of performance.
    2. Garrett will not use AI to build self-learning and self-adjusting security products that may perform at an adequate level when they are installed but then change over time through the influence of exogenous factors that are not prompted by the end user.
    3. Potentially sensitive data will not be collected by Garrett for our own development use without express client consent.
  7. Non-Product Development Uses of AI:
    1. Garrett uses AI in other business functions that do not relate to product development or the performance of our security products. The use of AI tools and methods in these applications are also governed by the Core Principles outlined in this policy and the other clauses which related to Data Privacy, Risk Management, and Compliance.
  8. Compliance:
    1. Garrett will ensure that our Artificial Intelligence Policy and implementation thereof aligns with relevant laws, regulations, and industry standards.
  9. Commitment:
    1. Garrett Metal Detectors believes that responsibly deployed AI enhances the capabilities of our products for security teams and treasure hunters alike. We keep safety, privacy, and trust at the core of everything we do.