Some Wyoming legislators are concerned that artificial intelligent bots, or perhaps even Chinese communists, could skew the outcome of the Bureau of Land Management’s controversial Rock Springs Resource Management Plan.
“It’s a little scary to think of a robot generating comments submitted under aliases and skewing things,” said Sen. Affie Ellis, R-Cheyenne, during a Friday meeting of the Select Federal Nature Resources Management Committee in Casper.
Rep. John Kolb, R-Rock Springs, said he was concerned that comments from people in “Communist China” or other parts of the world might carry as much weight as those submitted by his constituents as the BLM moves forward with the resource management plan (RMP).
No Specific Countermeasures
The BLM will focus primarily on the “substance” of comments submitted to the agency regarding the RMP, said Associate State Director Kris Kirby. That should give more weight to input from locals, and also help the BLM detect efforts to flood the comments site with fake identities or mass-generated submissions.
Even so, Kirby said she’s not aware of any specific countermeasures against generative AI spam, or methods of verifying the identity of everybody who summits a comment. So, that could be a legitimate concern.
“I agree. I don’t know if we have a good answer, but I think it’s something we need to follow up on internally and start talking about with our headquarters office,” Kirby said. “And maybe there’s things we’re talking about already that we’re not aware of.”
The Rock Springs BLM field office oversees roughly 3.6 million acres in Wyoming. The preferred alternative draft of the RMP designates 1.8 million acres of that as “areas of critical environmental concern” (ACES).
Please Submit Your Comments
The plan has stirred controversy since it was released in August. Many have said that the BLM’s preferred alternative is far too restrictive on energy development, cattle grazing and other uses that drive local economies.
And Gov. Mark Gordon’s office is considering suing the BLM over the proposal.
BLM officials have been adamant that public comments, particularly from Wyomingites who might be most effected by the RMP, are vital to the process.
Kirby reiterated that to the legislators Friday, saying that the on-the-ground perspectives of people who work and recreate in the Rock Sprigs Region could help shape the agency’s final decision.
The draft RMP, maps and a link to submit comments are available online.
Does The Whole World Get A Say?
However, Kolb said that the BLM will only give credence to what it considers “substantive” comments on the RMP, and what amounts to a substantive comment is up to the agency to decide.
Moreover, he balked at the idea that the BLM is taking comments not only from all over the country, but the entire world, including “Communist China.”
He asked Kirby if she thought that “the United States of America is still a sovereign nation,” and she responded “yes.”
“How on God’s green earth do you accept comments from the rest of the world as it’s associated with the RMP?” Kolb responded.
Kirby said that the detailed knowledge that people from Rock Springs and other parts of Wyoming will give makes their input carry more weight than those from elsewhere.
“(Local ) people are going to be able to focus, have that laser focus on specific pieces of the plan that they think might have impacts to their wellbeing, etc., and they can actually provide reasons they would and give us that detailed information,” she said. “Whereas, someone from another country or even another state — West Virginia or Georgia — might not be able to have that same articulation around an issue.”
Who’s Faking It?
That prompted committee member Rep. Reuben Tarver, R-Gillette, to ask whether the BLM can tell if somebody is faking being from Rock Springs when submitting a comment.
“I know people don’t lie typically in this world. But what if someone was to lie and pretend they were from
Rock Springs and they weren’t?” he asked Kirby.
Kirby reiterated that the agency will give more consideration to the detail and the substance of the comments than to where people might claim to be from.
However, she added that the agency doesn’t have any specific protocols for verifying commenters’ identities.
“That’s very concerning,” Tarver said in response. “We have some incredibly smart people. They’re wordsmiths. I’m not one of them, obviously, but that’s very concerning that we could have some outside influence. How would you know?”
Spam Bombs Won’t Work, BLM Says
Kirby and Brad Purdy, the BLM’s deputy state director of communications, said the agency is used to dealing with mass-generated comments such a “form letters.” So, AI spam would be handled the same way.
“If there’s a robot generating the same comment under multiple different names and putting it in 1,000 times, that’s only going to count as one comment,” Kirby said.
Purdy agreed that such efforts would be handled the same was as mass form letters from advocacy groups. Mass ChatGPT-generated comments might represent “creative thinking” on a commenter’s part, but would still be counted as only one comment, he said.
Ellis remarked that she’s concerned that artificial intelligence has advanced to the point that somebody clever enough could use it to generate comments that would at least appear to be coming from multiple sources.
“You could receive 10,000 comments from all over the country, from essentially one robot, or one person with an agenda,” she said.
Mark Heinz can be reached at firstname.lastname@example.org.