The Biden-Harris administration has announced that eight major tech companies have agreed to red team their AI applications and prioritize research that will make their systems more trustworthy.
The companies involved are Adobe, Cohere, IBM, Nvidia, Palantir, Salesforce, Scale AI, and Stability AI. They have all committed to developing machine-learning software and models in a safe, secure, and trustworthy way.
Under the agreement, the companies will submit their software to internal and external audits, where independent experts can attack the models to see how they can be misused.
They will also safeguard their intellectual property and make sure that the weights of their proprietary neural networks don’t leak.
Additionally, they will give users a way to easily report vulnerabilities or bugs, and publicly report their technology’s capabilities and limits, including fairness and biases, and define inappropriate use cases that are prohibited.
Also, they will focus on research to investigate societal and civil risks AI might pose if they lead to discriminatory decision-making or have weaknesses in data privacy.
The U.S. government is also encouraging the companies to develop watermarking techniques that can identify AI-generated content.
Finally, the companies have committed to building models for good, such as fighting climate change or improving healthcare.
The sources for this piece include an article in TheRegister.