Biden Admin. Seeks Public Input On Rules For AI Safety

Biden Admin. Seeks Public Input On Rules For AI Safety

Written By: Lauren Berg

Re-posted from Law360 (April 11, 2023, 10:07 PM EDT) — Artificial intelligence tools like OpenAI’s popular chatbot ChatGPT could offer a number of benefits, but the technology also raises safety concerns, the Biden administration said Tuesday, asking the public to weigh in on how the government can address the risks and create regulations to minimize potential harms.

The U.S. Department of Commerce‘s National Telecommunications and Information Administration asked the public for feedback on what kinds of policies the government can implement to make sure AI systems work as advertised without adverse consequences, according to the 31-page request for comment.

Commerce is interested in hearing about the kinds of safety testing AI developers should be conducting, what kinds of data access is necessary to conduct such assessments and what different approaches might be needed in different industries, like health care or employment, the request states.

“Responsible AI systems could bring enormous benefits, but only if we address their potential consequences and harms,” Alan Davidson, assistant secretary of commerce for communications and information and NTIA administrator, said in a statement Tuesday.

“For these systems to reach their full potential, companies and consumers need to be able to trust them,” he said. “Our inquiry will inform policies to support AI audits, risk and safety assessments, certifications, and other tools that can create earned trust in AI systems.”

But the Center for Data Innovation, a nonpartisan think tank that studies the intersection of data, technology and public policy, warned in a statement Tuesday that the alarm bells going off about AI systems could lead to rules that deviate from American innovation and competitiveness.

“Many of the proposals to regulate AI are based on the false premise that AI is fundamentally dangerous and consumers need the government to protect them,” Hodan Omaar, senior policy analyst at CDI, said.

Omaar argued that the government shouldn’t “bog down companies using algorithms with new regulations,” saying companies should instead be held accountable for monitoring their use of algorithms and mitigating potential harms. She said the Biden administration should enforce existing laws and regulations and conduct a gap analysis to pinpoint any shortcomings.

“If the administration allows the misguided notion that AI is inherently problematic to cloud the discussion on AI regulation, then the United States will become like Europe: aspiring to be an AI leader but in reality being an AI follower,” Omaar added.

The Biden administration’s request comes as the issue of generative AI — which has the ability to generate clear and human-sounding text about virtually any topic — heats up, with some watching its potential and others warning about the possible risks.

In particular, legal companies have been taking a close look at how attorneys could use AI-powered tools in their everyday practice.

Last month, the head of the Solicitors Regulation Authority told an audience of English lawyers that ChatGPT and other similar artificial intelligence chatbots will be “a game-changer” for the legal profession. Paul Philip, chief executive of the solicitors regulator, said that “technology is the next big-ticket item for the legal sector” and attorneys who ignore technologies like ChatGPT would do so at their peril.

Open AI released its free ChatGPT app at the end of November, allowing users to generate a written response that mimics a wide range of types of text. A million people had signed up within five days, reaching 100 million in January — making ChatGPT the fastest-growing web application in history.

The technology also has the potential to assist in some aspects of the laborious process of patent drafting, but attorneys told Law360 last month that it carries a risk of creating more work or torpedoing the patent before it’s even written. One attorney found that ChatGPT will produce text in a recognizable patent format but it doesn’t contain anything novel or non-obvious.

The Center for Artificial Intelligence and Digital Policy last month lodged a complaint with the Federal Trade Commission seeking to block OpenAI from continuing to commercially distribute its newly released software known as GPT-4, arguing that the product is “biased, deceptive” and fails to adhere to the commission’s standards for using such artificial intelligence.

And earlier this month, FTC member Alvaro Bedoya pushed back on the “myth” that there are no existing regulations that cover generative AI, saying the “unpredictability” of this emerging technology is no excuse for companies to ignore their obligations to protect consumers.

In October, the White House’s Office of Science and Technology Policy released a Blueprint for an AI Bill of Rights, which laid out a voluntary roadmap for the responsible use of AI.

The issue has also garnered international attention, with the European Union currently mulling a proposed Artificial Intelligence Act, which would create a risk-based framework to regulate AI applications, products and services.

–Additional reporting by Caroline Simson, Daniel Tay, Sarah Martinson, Marialuisa Taddia, Ryan Davis and Allison Grande. Editing by Dave Trumbore.

To read more go to: Click Here and get your Law 360 subscription today to view more great legal content.

If you or a loved one have been involved in a personal or product injury, or harmed by a defective products; contact Serna & Associates PLLC Today. Click Here to learn more or call: 1-877-822-1212 attorneys standing by.