In the days and hours leading up to Mark Zuckerberg’s testimony before the Senate Judiciary and Commerce committees on Tuesday, the consensus was clear: This was not going to go well. “The poor guy is going to be walking into an ambush,” The Cook Political Report’s Charlie Cook told NBC News. “I mean this is going to be like Custer’s Last Stand. I mean this is going to be Little Big Horn.” Writing in The Atlantic, Franklin Foer predicted a historic shift: “Tomorrow will be the day Zuckerberg raises his hand before Congress, and it will be the day Silicon Valley no longer floats above the world.” Axios even released a list of the senators who were most likely to eviscerate the Facebook CEO.

Hauling Zuckerberg before Congress was the culmination of a monthslong reevaluation of Big Tech in Washington D.C. on both sides of the aisle. His earlier media appearances to do damage control on the Cambridge Analytica scandal—in which tens of millions of Facebook users’ private data had been compromised—had been sweaty and awkward, with Zuckerberg clearly rattled by the severity of the backlash.

But after hours of testimony on Tuesday, Zuckerberg and his shareholders were likely breathing a sigh of relief. Little Big Horn this was not; rather, it was as if an errant yet gifted student had been hauled before the principal. With some exceptions, legislators kept their gloves on. Regulation, according to the majority of senators on the committees, would be an action of last resort. And if it was undertaken in the near term, it would be done with Facebook’s consultation. There was an implicit threat in the proceedings—if Facebook messed up again it would be in deep trouble!—but there was an abiding sense that, while a shift is underway in tech policy, Zuckerberg and his company were getting off with a warning for now.

Zuckerberg has a long history of skating past controversy. At Harvard, after he illicitly obtained photos of female students to build a program that would allow others to rank them by “hotness,” he was hauled before a disciplinary committee to explain himself. There he apologized—and chastised Harvard for its lack of digital security. He got off with limited repercussions.

“Frankly, Mark’s social awkwardness—and his confusion over the response to Facemash—had been his greatest defense,” writes Ben Mezrich in The Accidental Billionaires. “The gathered deans had looked at him and listened to his stilted affectation, and they had realized that Mark really wasn’t a bad kid—he just didn’t think the same way other kids did. He hadn’t realized that girls were going to get mad because guys were voting on their appearance.”

Zuckerberg did something similar on Tuesday. He apologized for breaking the trust of his company’s users and of the gathered senators. He emphasized again and again the humble origins of Facebook, which began in his dorm room in Harvard. He made the case that this was all a big misunderstanding—that Facebook was so committed to making the world a better place that it had temporarily lost focus on the ways in which its products could be used for harm. He stressed Facebook’s role in facilitating social change, mentioning #MeToo, March for Our Lives, and Hurricane Harvey relief in his opening statement. He stretched the truth, claiming that the company does not sell its data to advertisers when its entire platform is built for advertisers to use that data to target consumers. Above all, he said that this would not happen again.

Most of the assembled senators seemed to buy this line, in part because many of them didn’t quite seem to understand what Facebook actually does. Between that and the limited amount of time they were given for questioning, the hearings at times had the feel of Zuckerberg explaining the internet to a retirement home. They were confused about Messenger and Instagram, largely ignorant of research suggesting that Facebook is addictive and makes people unhappy. Most problematically, they failed to grasp that its business model basically guaranteed that large-scale invasions of privacy would become the new normal.

That ignorance led to some embarrassing missed opportunities, such as when Senator Bill Nelson of Florida used the example of chocolate—as opposed to, say, a more personal product like underwear—to discuss the company’s rampant use of ad retargeting, in which a user searches for a product online, only to see ads for that product follow them from platform to platform and site to site. Zuckerberg’s response was telling, even if it was the company line: There may be “some discomfort” about ads following you across platforms, he said. But “the overwhelming feedback we get from our community is that people would rather us show relevant content than not.” This, in a nutshell, is Facebook’s line on privacy—that people actually like having it violated.

Faced with tougher questions, Zuckerberg did two things again and again. He made bold claims that artificial intelligence would be able to block misinformation and hate speech—and cure other ills that have afflicted the social network—within “five to ten years.” And, pressed on specifics, he told skeptical senators that he would have someone get back to them—a response that most found more than satisfactory.

Given the general lack of understanding about Facebook and technology that was on display, it should come as no surprise that even the senators who pressed Zuckerberg the hardest, like Lindsey Graham, made it clear that they weren’t prepared to offer regulatory action without the company’s participation. Again and again, legislators asked Zuckerberg how he would like them to act, as if they were working in partnership. It’s hard to imagine that any meaningful action can take place in an environment where senators are so ignorant of Facebook that they basically begged Zuckerberg for his help in crafting regulation—during a hearing where they were supposed to be grilling him.

But even if Zuckerberg’s testimony didn’t necessarily usher in an immediate, drastic change in Big Tech’s relationship with Washington, there were ample signs that a subtler development is afoot. Graham pushed Zuckerberg hard on the question of whether or not Facebook is a monopoly (it is), an example of the growing bipartisan push for stronger antitrust action. Senator Dick Durbin, who asked Zuckerberg if he would tell the gathered senators where he stayed last night, got the Facebook CEO to admit in a public, embarrassing way that he valued his privacy, while his company devalues that of its users. Senator Patrick Leahy pushed Zuckerberg repeatedly on Facebook’s alleged role in propagating the Rohingya genocide in Myanmar.

Zuckerberg is hoping that this apology tour, like the many that preceded it, will ultimately end with Facebook returning to business as usual. But over the last two weeks, Zuckerberg has laid out a set of new criteria by which his creation should be judged by Congress. While his commitment to new regulations comes with a giant asterisk—he always emphasizes that he is amenable to the “right regulations,” which basically means “minimal changes”—the door is clearly open in a way that it wasn’t before.

“It’s not enough to just build tools,” he said early in the hearing. “We need to make sure that they’ll be used for good.” That’s a promise that Facebook can’t live up to just yet, and Zuckerberg surely knows it. The Federal Trade Commission and some states are beginning to test that premise. Congress, as we saw on Tuesday, has been slow to catch on, but there are signs that this is changing. And the next time Facebook violates its users’ trust—as it inevitably will, if history is any guide—Zuckerberg might get a reprimand tougher than the equivalent of detention.