The National Telecommunications and Information Administration (NTIA) has issued a recent policy report aimed at holding AI developers accountable for their creations. Anyone who relies on AI for videos, articles, art, even decision making, can fall under this new framework. This move marks a significant shift in the regulatory environment and has profound implications for QA processes and risks for AI companies.
The Essence of NTIA’s AI Accountability Rules
The NTIA’s policy report, released on March 27, 2024, defines touch points for AI accountability. A lot of the language will sound familiar if you have read the EU AI Act or the National Institute of Standards recommendations. Security, transparency and being unbiased or equitable are the key requirements for AI developers. Also, the report provides guidance, support and standards for creating trustworthy AI.
Implications for AI Companies
For AI companies, the NTIA’s rules signify a new era of responsibility on top of the recent spat of U.N., EU and U.S. state rules pouring forth from legislatures. The policy stipulates that developers and deployers of AI systems must assure their offerings are safe and reliable. In practice, companies now need to implement thorough and in situ evaluation mechanisms through the development cycle. The end goal is to ensure the AI systems are tailored for the targeted purposes and as risk free as possible.
- Embracing Transparency and Trust
AI companies will be expected to disclose information about their systems’ capabilities and limitations. Users are expected to be able to make decisions about whether they can trust the AI system based on the level of transparency about the training and the model. NTIA anticipates that this transparency and trust rating will become a competitive marker in the industry, much like independent reviews. - Prioritizing Ethical Considerations
Every AI developer and consumer facing user will need to expressly define how potential biases and questionable ethics were removed from models. Rather than waiting for consumer complaints regarding unethical or biased results, AI companies will need to either subcontract out monitoring or build in processes to ensure the ethical, unbiased or non-disriminatory results the new framework requires. - Adapting to Regulatory Requirements
AI audits and control systems (this is what DSG can take care of for you:)) are required per the NTIA’s policy. The easiest and most professional way to adhere to this requirement is to subcontract out the process to a company like DSG since keeping up with the latest rules keeps the responsibility off of your company’s shoulders. - Facing Consequences for Non-Compliance
NTIA’s policy states there will be financial and legal consequences for not adhering to the standards. Reputational damage could well be as significant, or even moreso, than the financial cost. Avoiding both the financial penalties and legal consequences is business critical for AI companies to focus on their core value add. DSG can help keep the regulatory wolves at bay by handling the monitoring and control aspect of AI implementations. - Fostering Innovation within Boundaries
Even though the policy sets boundaries for AI development, it also encourages innovation. The premise is that by offering AI models that are transparent, ethical, unbiased, and targeted, the industry will be able to innovate more rapidly with full market support.
Conclusion: A Call to Action for AI Companies
Some AI companies will need to reassess their approach to AI development. The call to action is that companies must now balance innovation with accountability, and confirm that their AI systems are not only advanced but also aligned with current societal values and safety standards.
As professionals in AI, we will adapt to these guidelines and work towards creating AI that earns the trust of users and stakeholders alike. The NTIA’s policy is more than a set of rules, it’s a blueprint for building a future where AI enriches our lives in line with values or never abridges our safety.
DSG.AI is your resource for keeping your AI development or application on track with the ever-changing regulatory environment.
Contact us today for a free consultation on how our services will add value to your AI.
