Teresa McManus, Local Journalism Initiative Reporter
The City of New Westminster has established an AI Working Group to address the rapidly evolving world of artificial intelligence.
City Council recently received a report outlining the work staff have done to date in developing appropriate AI usage guidelines and future work.
The July 13 report said public-facing generative AI (GenAI) platforms have seen a “surge” in popularity since their arrival in late 2022 and 2023. He said more than 15,000 GenAI platforms and tools are currently available online, with new tools being announced every day.
“Currently, the city does not have specific tools that would enable meaningful and detailed reporting on the use of the GenAI platform by city employees, contractors, volunteers, and students,” the report states. “Advanced insights have alerted IT that some staff are using corporate devices to leverage various GenAI tools, but detailed reporting is not possible at this time.”
Staff say many internet search platforms incorporate extensive language models and direct conversational responses into their standard search plan forms, making it difficult to avoid interacting with certain AI prompts.
“This trend will only increase as Google and other organizations further integrate advanced AI models into traditional web search,” the report said. “In addition to online AI platforms, software vendors are increasingly building AI capabilities into applications installed locally on workstation computers, such as the Adobe suite.”
The report highlights some of the “numerous risks” associated with the use of generative AI models. These include:
- Data privacy concerns
- Compliance and regulatory violations
- Misinformation and disinformation
- Impact on cybersecurity
- AI bias and discrimination (generative AI models learn from the data they are trained on)
To reduce the risks associated with AI, city officials have developed and published a set of guidelines for employees, contractors, volunteers, and anyone working on behalf of the city, including the mayor and city council. These guidelines are designed to:
- Increase productivity by using GenAI tools safely and in ways that benefit your community
- Ensure compliance with all applicable policies and laws
- Protecting personal or confidential information stored and managed by the city
- Reduce reputational, legal and operational risks
“While the guidelines are an important first step, more work is needed to improve awareness and understanding of risks and how to mitigate them, ensure compliance, and protect City assets from the various risks identified,” the report states. “As part of this effort, our IT department has been researching a variety of software tools that provide a high level of protection against these risks and also provide detailed usage reporting.”
county. Nadine Nakagawa asked whether it would be considered a violation of nonpublic information if a lawmaker or staff member inputs information from private meetings or reports into an AI model.
City Finance Director Natalie Fröhlich confirmed this would be a violation under city guidelines.
Nakagawa said it’s “very important” that city officials are aware of this because people use AI for a variety of purposes.
Nakagawa said Generative AI “has a huge impact on the climate and the environment,” and that he hopes the working group takes this into account as it continues its work.
“I think there is good guidance for staff, for example, that it is a supplement to judgment rather than a substitute for it,” she said. “I would also like people to think about whether it is necessary given the huge amount of water usage and electricity usage. There is no doubt that people are using this, and I think using it responsibly means recognizing that every time you use generative AI, you are consuming resources, even if you don’t see it.”
Citing issues such as accuracy, professional judgment and originality, Coun. Tasha Henderson said she believed it was important that councilors’ comments to residents were in their own words and not written by an AI.
“I think as elected officials, we are expected to share our own ideas,” she said. “So I now have a disclaimer in my[email]signature to let residents know that I never use AI in the emails I respond to, in my social media, in my floor remarks, etc. And I would encourage similar transparency from my colleagues here on Congress in terms of letting people know when and when not to use AI in my comments as a councilor.”
Mayor Patrick Johnston said he has not personally taken the time to use AI tools, but he does use various online programs such as Word, Adobe and Excel. He acknowledged that council members should not save information from private meetings on computers that do not belong to the city, but he suggested that council may need more training given changes in technology.
Johnston suggested that staff might want to consider providing training on AI to councils, perhaps as part of council training after the fall election.
“I think we’re moving into an area where it’s very difficult for a lot of people to understand where the boundaries are.”
