This Week's Buzz in Silicon Valley: Agentforce dazzles, but AI ethics loom

Agentforce was a big success for Dreamforce and Salesforce, no doubt about it. And, San Francisco, which has its fair share of problems with homelessness and drugs, seemed to have things under control this time, much to everyone’s relief. But of course, a tech event like this is never without a few bumps.

At Dreamforce, California Governor Gavin Newsom introduced his new anti-deepfake law, showing a deep fake video of Kamala Harris, a fellow Democrat and presidential candidate, as an example. Salesforce CEO Marc Benioff quickly showed his support, but that didn’t stop the flood of criticism on social media platforms—accusations of populism, political bias, and threats to free speech.

Elon Musk calls the new law “an attack on free speech” and has shared the video of Harris on Twitter with the following text: "The governor of California just made this parody video illegal in violation of the Constitution of the United States."

Others are worried that the law could open the door to censorship during elections. The fear is that satire and jokes could fall under the broad definition of deep fakes, raising concerns from legal experts and civil liberties groups.

The law requires social media platforms to moderate or label deep fake content during elections. It’s the first law of its kind to impose this responsibility on platforms, and it’s expected to impact the 2024 presidential race. However, big tech companies will likely challenge it in court, arguing it limits free speech.

The law is already facing legal battles, with opponents claiming it violates the First Amendment. These legal fights could last for months, maybe years. The outcome could set an example for other states or see the law struck down before it takes effect.

Another talk piece at Dreamforce was Disneyland’s decision to stop using Slack by 2025. This move comes after a massive data breach in July 2024, when the hacker group Nullbulge leaked over 1.1 terabytes of sensitive data from Disney’s Slack channels. This included project details, financial information, and internal discussions about Disneyland Paris. Disney sent an internal memo announcing the shift to Microsoft Teams, though they didn’t directly mention the breach. However, the timing makes it clear the data leak was a significant factor. 

Marc Benioff declared that he was a true Disney lover during the Agentforce launch and then defended Slack’s security. According to Benioff, human error is often to blame in breaches. Still, the incident reminded us of how vulnerable even the most widely used communication platforms can be, and it took some of the shine off the event.

This year, 300 people from the Nordic countries flew in to attend Dreamforce, including representatives from the Norwegian Consulate in San Francisco and several prominent figures from the Nordic community in the Bay Area. 

With Newsom’s new laws and ongoing debates about AI manipulation, the real question is no longer whether we can trust what we see—but how to protect future elections and debates from being overwhelmed by AI-driven falsehoods without threatening free speech or turning the issue into a political battle. This is important not only for California but for the whole world.