Artificial intelligence tools already assist with tasks up and down the legal workflow if attorneys are savvy enough to invest in them. But the technology is still a long way from reliably performing the sort of “robot attorney” role that everyone either dreams of or deeply fears. For now, the best AI is one that understands (or is made to understand) its own limitations being used by conscientious attorneys focused on keeping AI from making stupid mistakes that compound into real world problems.
We’re putting a lot of faith in those conscientious attorneys despite living in a world where the opposite hold far too much sway. The sort of folks who would just throw AI at a problem it’s uniquely ill-equipped to solve and pocket the fees while ignoring the fallout.
Did you know the United States government uses AI to translate refugee asylum requests? Because it does and it’s generating the sort of human tragedy you’d expect from that sentence:
A crisis translator specializing in Afghan languages, Mirkhail was working with a Pashto-speaking refugee who had fled Afghanistan. A U.S. court had denied the refugee’s asylum bid because her written application didn’t match the story told in the initial interviews.
In the interviews, the refugee had first maintained that she’d made it through one particular event alone, but the written statement seemed to reference other people with her at the time — a discrepancy large enough for a judge to reject her asylum claim.
After Mirkhail went over the documents, she saw what had gone wrong: An automated translation tool had swapped the “I” pronouns in the woman’s statement to “we.”
You might be wondering how the hell the COURT failed to untangle this issue itself, but that would put a lot of unearned trust in a government process where they contend 3-year-olds can competently represent themselves. We’re not exactly dealing with an entity that puts a lot of energy into protecting the refugee’s rights.
How are these automated translators in charge of such a critical task? It’s a lot cheaper. U.S. Citizenship and Immigration Services uses Google Translate in reviewing asylum applications. As the Rest of the World reports, “Major translation companies like LanguageLine, TransPerfect, and Lionbridge have contracts with U.S. federal immigration agencies, some totaling millions of dollars.” Having turned over critical government functions to private contractors, we’re stuck with those contractors seeking every efficiency to maximize profit and in cases like these the contractors can lean on AI and expect the now overworked and potentially disinterested government client to be responsible for checking its work.
This is the how AI is going to make things worse before it makes them better. For every well-financed and responsible firm investing in this tech, there are going to be government agencies struggling against lean budgets and politically driven “privatization” edicts that buy unfiltered hallucination machines and turn them loose on serious matters.
And it’s going to disproportionately be serious matters that impact the least powerful.
AI translation is jeopardizing Afghan asylum claims [Rest of the World]
Earlier: The Legal Industry Has A Long Way To Go Before GPT Matches The Talk
DOJ Official: 3-Year-Olds Can Competently Represent Themselves
Joe Patrice is a senior editor at Above the Law and co-host of Thinking Like A Lawyer. Feel free to email any tips, questions, or comments. Follow him on Twitter if you’re interested in law, politics, and a healthy dose of college sports news. Joe also serves as a Managing Director at RPN Executive Search.
Leave a Reply