Google has revealed a series of measures aimed to better inform US voters while reducing the risk of malicious misinformation. It includes restrictions on its AI tools and better protection against cyber attacks for election and campaign officials.
The measures broadly break down into three categories. The first is better information for users who are trying to find out more about elections. For example, users who search YouTube for the name of an election candidate will get an information panel detailing official website and video channel links for the candidate. (Source: engadget.com)
Similarly, people searching for voter registration will get a list of “authoritative information” from state election offices. This will effectively be an override of whatever the normal Google algorithm would consider the “best” results for such a search.
The second category is artificial intelligence, something that’s timely given Google’s recent heavy push for its Gemini tool. Results created by Gemini will be heavily restricted or removed from search results or video summaries relating to election topic. Again, that’s something of an admission that the results aren’t reliable or trustworthy enough for such a sensitive topic.
Google is also reviewing the ways people may use AI and similar tools to try to violate some of its core policies on elections such as when it sells online advertising. For example, while candidates can make claims about their rivals that some may consider misleading, Google doesn’t automatically crack down on this. Contrastingly, advertisers cannot make false claims about voting dates to mislead opposing supporters.
Google notes “we are not seeing dramatic shifts in tactics compared to previous election cycles. To the extent we’re seeing bad actors use AI, it is generally to increase the speed or scale of their information operations, rather than deploying new strategies.” (Source: blog.google)
Finally, Google says its continuing to offer strengthened account security and defenses against denial of service attacks and other cyber threats against users at high risk during an election period. As well as election officials and candidates, this includes journalists. The theory here is that a hijacked journalist account could be used to spread false information more credibly.
Are these smart moves by Google? Do any of them go too far? How much responsibility do big tech firms have for reducing misinformation during an election period?
