VISTA — The rapid growth of artificial intelligence, or AI, is one of today’s hottest topics.
From professors trying to weed out AI-written assignments to intelligence models “escaping” their immediate environments, seemingly no corner of society is immune to the turmoil stirred up by AI’s growing influence.
To get ahead of potential issues associated with the technology, the Vista Irrigation District recently adopted an official AI policy, becoming one of the first local agencies to do so.
Shallako Goodrick, Vista Irrigation District’s director of administration, said the policy came about at the board’s suggestion.
“We’re not usually on the cutting edge of things,” Goodrick said. “It really came about because our directors go to a lot of conferences… One of our board members, who reads a lot of news, saw that the city of San Diego had created a sort of AI coalition.
“The board asked us to look into it. We don’t use AI here currently. Employees might use free AI (tools) to help them write stuff. But we’ve never deployed AI and have no intention to do so anytime soon. The board still thought we should have an overarching policy.”
The board adopted the proposed policy at its Sept. 9 meeting.
“Ours is very high-level: ‘We will be responsible.’
According to Goodrick, one of the biggest issues is confidentiality and discrimination in AI models.
Once the decision to craft an AI policy was made, Goodrick said she began looking at what other government agencies had already done.
“I try not to start from scratch if I can. I definitely used San Jose’s AI coalition. They had a template that agencies can use, so I used a lot of pieces of that. I used a few others that I compared it to,” Goodrick said. “The other one I used is called the Municipal Information System Association of California. I hadn’t heard of them before. They were very similar.”
When asked the obvious question, Goodrick laughed and said, “I did not use AI to write our AI policy! I’m not a big user of AI yet.”
While the district does not currently use AI in its operations — Goodrick said its IT team even tried unsuccessfully to disable the built-in AI model embedded in the office productivity software it uses — Vista Irrigation leaders recognize that employees may use AI to assist with their work, particularly when writing or editing reports.
“A big piece of it is educating staff on how can you safely use AI to help you with certain things, but still retain responsibility. With any tool, you’re ultimately responsible for the outcome. Part of this is internal training. Employees need to be trained on the limitations of AI, and validation.”
Another concern for the water district is how its vendors might use AI in ways that could affect VID customers.
“The next thing is adding AI in our contracts, if vendors are using it. We’re kind of at the beginning of all this and trying to find our way. It would be similar to, ‘You have to make sure your AI is not discriminatory, and have control of it.’ ”
“So far, I’ve only seen our vendors using it in chatbots. Everybody sees it there – and that’s where I’m seeing it with our vendors.”
Goodrick said the vendor whose use of AI would most likely affect VID customers is the third-party company the district uses to process credit card payments.
With AI evolving so rapidly, how often will the district revisit its policy?
“We tend to review most of our policies on an as-needed basis. Obviously, we have our eyes on AI. For the most part, unless we went to deploy our own AI I don’t know what changes would take place.”
And it’s not just advances in AI technology that could prompt changes to the policy.
“Usually what causes us to review our policies really closely is laws get passed.”
