(From left): Arlene Zalayet, GC and SVP at Liberty Mutual Group; Aryeh Friedman, chief privacy officer and associate GC at Dun & Bradstreet; Charles Neff, director of compliance at Huntington Ingalls Industries; and Carolyn Metnick, a partner at Ackerman LLP. Photo: Dan Clark/ALM

AUSTIN—While artificial intelligence (AI) is the hot topic in in-house legal departments in 2018, the ethics governing its use are not fully realized. A panel of attorneys, both in-house and in private practice, however, said on Monday that the rules of professional conduct are perhaps the best way to approach the use of AI.

Carolyn Metnick, a partner at Akerman LLP, who is co-chair of its privacy, cybersecurity and emerging technologies team based in Chicago, said that it may still be a couple of years before guidance is issued on artificial intelligence.

Metnick spoke at a panel on artificial intelligence at the Association of Corporate Counsel annual meeting in Austin at the Austin Convention Center.

“The bar associations have not put out much guidance on artificial intelligence,” she said.

The best way to guide yourself is to look to four Rules of Professional Conduct, Metnick said. Specifically, Rule 1.1, the duty of competence; Rule 1.4, the duty to communicate; Rule 1.6, the duty of confidentiality; and Rule 5.3, the duty to supervise, she said.

Aryeh Friedman, the associate general counsel and chief privacy officer of Dun & Bradstreet, agreed that the duty of competence is key. He said that having a legal department made up only of only lawyers can no longer be, and that technologists are necessary to help lawyers understand artificial intelligence so that those lawyers can explain the machine's findings.

Charles Neff, director of compliance at Huntington Ingalls Industries, a shipbuilding company, also said it is important to know the risks. Knowledge of those risks would fall under the duty of competence. The panel agreed.

“You need to be able to explain it to the client,” Metnick said.

Neff said among those risks that need to be explained are that bias is inherent in AI and machine-learning tools. He cautioned that those who use it must understand the necessity to filter certain results.

Metnick also explained that AI requires the duty of confidentiality. Often third-party vendors use AI and it is important to get the client's consent to allow those vendors to have access to confidential information, she said.

“If you're going to be working with an outside vendor, you should discuss privacy with that vendor and see what security measures are put in place,” Metnick said.

Finally, those using AI should also adhere to the duty to supervise.

“We have an obligation to supervise the work of lawyers and nonlawyers and to make sure we are all following the rules of professional conduct,” Metnick said.

The panel was moderated by Arlene Zalayet, the senior vice president and general attorney of Liberty Mutual Group.