Axios

NEW YORK – AI and the internet are outpacing oversight — threatening kids, creatives, national security and even basic innovation, leaders across tech, politics and entertainment said at the Axios' AI+ Summit.

Why it matters: AI is transforming industries and society faster than it can be regulated, creating sweeping, serious security risks, according to several speakers.

The June 4 summit hosted multiple conversations and was sponsored by BCG, Booking Holdings, Snyk, Varonis, and Workato.

Here are some key takeaways:

• WndrCo founding partner Jeffrey Katzenberg said kids' unsupervised use of the internet is "destroying a generation."
• Lumen Technologies president and CEO Kate Johnson said telcos aren't innovative enough and have ceded too much ground to Big Tech.
• The Weather Company CEO Rohit Agarwal said AI could help forecasters be as specific as giving guidance on what time of day to walk your dog.
• Gov. Kathy Hochul (D-N.Y.) made a dig at Rep. Marjorie Taylor Greene (R-Ga.) for saying she didn't know the GOP's "big, beautiful" tax bill included a provision that would ban states and municipalities from regulating the tech for 10 years.
• Lux Capital co-founder Josh Wolfe said the best way to beat China in the AI race is to "make sure every single young" person is super well-versed in AI.
• Actor and entrepreneur Joseph Gordon-Levitt said there needs to be an incentive to keep creatives paid and employed as AI disrupts the entertainment business.

Content from the sponsored View from the Top conversations:

Vlad Lukić, BCG managing director and senior partner, and global leader for its tech & digital advantage practice discussed the disconnect between corporate AI investment and tangible outcomes.

• According to a recent study of 1,000 companies, "Over 75% of them are with budgets in this year deploying AI at scale, but only 25% of them have a line of sight to value creation from those activities," he added.

Danny Allan, chief technology officer at Snyk, said visibility, false expectations, and proper policies are lacking behind the pace of AI-powered software development.

• "The speed and velocity that software is coming through the pipelines is like nothing I have ever seen in my career right now. It's so, so fast. And the trouble that CISOs have is they don't have the trust that what is coming through that pipeline is actually secure."

Bhaskar Roy, Workato chief of AI products and solutions, said businesses will experience real transformation when AI agents tackle the "messy middle."

• "There are a few companies that are targeting the core and looking at how they can transform the core with … agentic AI and that's what excites us."

Rob Sobers, chief marketing officer at Varonis, warned that security risks, like "AI model poisoning" where attackers inject malicious data into AI models, could impact people's lives.

• While working with an organization researching Alzheimer's, they noticed a hacker feeding the organization's custom AI model new data out of nowhere, Sobers said. And, "that could change subtly the dosage of a medication that you're giving somebody. … It's super important to get the trust and security layer right."

United Natural Foods Inc., the primary distributor for Whole Foods, said Monday it had to take some of its systems offline after an apparent cyberattack.

Why it matters: In the short-term, the company might not be able to fully fulfill customers' orders as it investigates the potential hack.

Apple on Monday announced that developers would be able to use Apple Intelligence models for their own apps, even as critics worry the company is falling further behind in AI.

Why it matters: Apple has been the slowest of the major tech companies to incorporate generative AI, with some features from last year still unreleased.

The wildest, scariest, indisputable truth about AI's large language models is that the companies building them don't know exactly why or how they work.

• Sit with that for a moment. The most powerful companies, racing to build the most powerful superhuman intelligence capabilities — ones they readily admit occasionally go rogue to make things up, or even threaten their users — don't know why their machines do what they do.

The Democratic National Committee is launching its first-ever live, daily show on YouTube, Axios has learned.

Why it matters: The party has been reeling since its losses in the 2024 elections, but it's counting on growing discontent with President Trump to help fuel interest in the show.

Few tears will be shed in Silicon Valley or at Big Tech firms over Elon Musk's precipitous fall from White House grace.

Why it matters: Musk's brief alliance with President Trump warped the usual dynamics of the relationship between America's most valuable industry and its center of political power.

President Trump reportedly picks up when his cell rings even if he doesn't know who's calling. Senior members of his team also love chatting on their personal devices.

• That makes the administration uniquely vulnerable to basic scams like spoofed calls and impersonation attempts.

Why it matters: If Trump is willing to answer unknown numbers, as The Atlantic reported this week, there's no guarantee a scammer, impersonator, or even a foreign intelligence operative couldn't have a chat with the president.