
Haugen questions whether tech giants will exploit loopholes in recent White House AI agreements
Whistleblower Frances Haugen warns AI companies to adhere to the spirit of their recent White House self-regulation agreement, expressing skepticism that firms will avoid exploiting loopholes despite recent proactive engagement with safety concerns.
AI-generated summary
Frances Haugen leaked internal documents in 2021 revealing Facebook's awareness of its platforms' negative impact on teen mental health. The issue of AI safety has recently gained significant attention from both regulators and industry leaders.
Facebook whistleblower Frances Haugen on Friday said artificial intelligence companies must "step up and comply" with the spirit of their White House agreement to self-regulate, days after Meta's Mark Zuckerberg and other CEOs signed the document.
"The history of tech companies engaging with regulation is: Often they will follow the exact written statement when we are dealing with very flexible systems," Haugen told CNBC's "Squawk Box" in an interview Friday.
She questioned if the companies will hold to the "spirit of the rules," or if they will seek to exploit things that aren't expressly prohibited and "go around the end of the fence."
Haugen, a former product manager on Facebook's civic misinformation team, leaked a trove of internal documents to the Wall Street Journal in 2021 that became a series of reports called "The Facebook Files." Some of the documents showed the company was aware of how its platforms negatively impacted teens' mental health.
"The Social Reckoning," a new Aaron Sorkin movie releasing Oct. 9, centers on that story. Haugen also wrote a book about her role in bringing the documents to light, "The Power of One: How I Found the Strength to Tell the Truth and Why I Blew the Whistle on Facebook."
As the debate around AI's threats to the human race has reached a fever pitch in recent weeks, Haugen said AI companies have been "more proactive" at engaging with safety concerns today than in years past. She also noted the push from Anthropic and its CEO, Dario Amodei, to advocate for independent auditors.
"I think that is an acknowledgement that, that at least at least Dario knows that that if you're given the space to cut corners, even the angels among us begin to cut corners," Haugen said.

Circuit Breaker Labs, a TechCrunch 2026 Startup Battlefield finalist, uses AI agents simulating diverse users to test models for psychological safety risks, motivated by lawsuits linking chatbots to youth suicides. The startup aims to prevent AI from enabling harmful parasocial relationships through adversarial testing and explainable scoring.

According to Bloomberg's Mark Gurman, Apple is preparing to officially introduce MacBook models with 14-inch and 16-inch screen options, touch OLED screens and 5G supported M5 Pro and M5 Max processors in late October or November. The new models promise significant improvements in color accuracy and energy efficiency by switching to OLED panels in display technology. Additionally, the adaptation of the Dynamic Island interface to laptops and 5G connection support will change the user experience and cloud-based work processes.

Creators who previously avoided Substack's 10% transaction fee are now facing a new price increase of at least $10 per month, according to The Verge Daily.

FELG Software reported a possible leak of patient data from the system for dental offices, which could have affected up to 2.4 million records. The incident is being monitored by CSIRT CeZ and other authorities, and the company ensures that the system remains safe to use. This is another attack on the health sector in Poland in recent weeks.

The “Global Digital Trust Insights 2027” report issued by PwC revealed that organizations are still far from reaching a unified model for distributing responsibility for artificial intelligence risks, despite the increase in spending on cybersecurity and the increased use of artificial intelligence in operations, as 29% believe that the responsibility falls on the responsibility of the IT leader, and 26% see it as the responsibility of an independent department for artificial intelligence, while artificial intelligence has become an increasing target for attacks, with half of security leaders indicating a lack of preparedness to confront it.

The new site of the Luiss Guido Carli University won first place in the Education category of the International Splash Awards 2026, international recognition for the best Drupal projects. The award, presented in Rotterdam during DrupalCon Europe, rewards fifteen months of work that rationalized content from 65,000 to 25,000 pages, reduced publication times by 70% and increased performance tenfold, guaranteeing an uptime of 99.9%. The project includes a multi-brand Design System and conversational search based on Artificial Intelligence.