Conspiracy theories are unavoidable following the Trump shooting. But social media firms are making things worse

Former President Donald Trump after being shot during a rally on July 13, 2024 in Butler, Penn.

The truth just had a real bad weekend.

Social media has long played a major role in making society more conspiracy-minded, by creating filter bubbles and stimulating the virality of outrageous notions to serve a business model built on the never-ending quest for engagement. The result is what we’ve seen online following the shooting of former President Donald Trump on Saturday.

On Facebook, X and elsewhere, falsehoods about what happened quickly became ubiquitous. Some were straight-up kneejerk partisan conspiracy theories, such as the dueling claims that the assassination attempt was staged by Trump’s team or ordered by President Joe Biden. Some seem to be fragments of early reports that quickly turned out to be untrue but that live on regardless, like the idea that Trump was wounded by a piece of teleprompter glass rather than a bullet.

Some conspiracy theories were propagated not merely by social media, but by social media CEO Elon Musk himself, who used X to suggest that the Secret Service might have deliberately ignored Thomas Matthew Crooks as he took aim from a nearby roof (85 million views). Musk also agreed with a post alleging that the shooter would “almost certainly” turn out to be Antifa (207,000 views. Crooks turned out to be a registered Republican.)

(Also of note: Musk used the occasion to endorse Trump’s presidential bid and reveal that two people had “tried to kill me” in the last eight months.)

In an information whirlwind like this one, it would be incredibly hard to stop conspiracy theories flying across social media even if the proprietor wasn’t enthusiastically joining in. The situation isn’t helped by the fact that—certainly at the time of writing—we still don’t know what motivated Crooks.

But the social media platforms have also made decisions that exacerbate situations such as this.

In a move that was likely inspired by Musk’s free-speech-first stance at X, Meta last year diminished the impact of its fact-checking program by allowing Facebook and Instagram users in the U.S. to stop suppressing debunked posts in the feeds they see. YouTube allows 2020 election denialism these days. Meanwhile, X’s algorithm is shoving conspiracy theories about the assassination attempt into its trending topics feeds, and Musk’s decision to change the meaning of a blue check from “person who knows what they’re talking about” to “person who pays to appear authentic” was always going to turbo-charge the spread of disinformation.

The deceptive nature of X’s blue-check system has now earned the company preliminary charges under the EU’s new Digital Services Act (DSA), which covers content on major online platforms. “There is evidence of motivated malicious actors abusing the ‘verified account’ to deceive users,” the European Commission, which has been formally investigating X since December, said Friday.

Musk responded with the assertion that the Commission “offered X an illegal secret deal: If we quietly censored speech without telling anyone, they would not fine us. The other platforms accepted that deal. X did not.”

All three of the EU charges against the company are about its lack of transparency; none is about censoring content. But there is also no great mystery about the fact that the DSA requires large platforms to fight disinformation and take down illegal content. These responsibilities are written into the law, which Musk explicitly endorsed two years ago while standing next to Commission digital chief Thierry Breton. “There has never been—and will never be—any ‘secret deal’. With anyone,” Breton tweeted Friday in response to a legal threat from Musk. “See you (in court or not).”

I am extraordinarily cautious about offering advice that touches on the U.S.’s febrile political moment, not least because I’m not American and I live on the other side of the world. However, I will say this: Given how systemically broken social media is when it comes to the spread of disinformation, it’s really up to each individual to be extra-careful about the ideas they spread and skeptical about what they read, especially if it fits too neatly with their own outlook. The truth is out there, but it has strong competition.

More news below—and do check out the livestream for Fortune Brainstorm Tech, which begins this afternoon in Park City, Utah. You might particularly want to tune in for a session with cybersecurity startup Wiz at 3:15 p.m. MT, for reasons that will become clear below.

David Meyer

Want to send thoughts or suggestions to Data Sheet? Drop a line here.

NEWSWORTHY

Alphabet eyes Wiz. Alphabet is reportedly considering making its biggest acquisition yet. The Wall Street Journal says the Google parent is in advanced talks to shell out $23 billion for cloud cybersecurity firm Wiz, which was earlier this year valued at $12 billion in a funding round. As the paper notes, this was a rare up round in a year when it’s generally only AI companies that get to raise cash at an increased valuation.

OpenAI whistleblowers. OpenAI illegally forbade its workers from warning regulators about the safety risks of its technology, company whistleblowers have told the Securities and Exchange Commission. “I don’t think that AI companies can build technology that is safe and in the public interest if they shield themselves from scrutiny and dissent,” one whistleblower told the Washington Post—which also reported separately that OpenAI last year forced its staff to race through safety testing of its GPT-4 Omni model in order to meet a predetermined May launch date.

Intel chip concerns. A few months ago, Intel started investigating increasing reports that its latest high-end i9 PC processors were causing games to crash. Now, according to The Verge, some servers using the i9-13900K and i9-14900K chips are also crashing games, leading Path of Titans maker Alderon Games to switch to AMD for its game servers, and recommend that other hosts of its game follow suit.

SIGNIFICANT FIGURES

$300,000+

—The amount AT&T paid a hacker to delete stolen data on almost all its customers, according to a Wired report. The hacker, from the ShinyHunters group, reportedly wanted $1 million before settling for around a third of that.

IN CASE YOU MISSED IT

Tech’s war for AI talent is creating a ‘massive consolidation’ problem, lawmakers say, by the Associated Press

Apple stock poised to rise 11% on iPhone AI updates, says BoA, by Niamh Rowe

A $3 billion software unicorn scrapped its plan to give AI ‘workers’ rights after tech execs said it ‘disrespects the humanity of your real employees’, by Sasha Rogelberg

Tesla’s AI dreams may ‘materialize on a longer time horizon (or not at all),’ UBS warns, by Will Daniel

California AI bill SB-1047 sparks fierce debate, Senator likens it to ‘Jets vs. Sharks’ feud, by Sharon Goldman

EV maker VinFast slashes full-year sales target by 20% as U.S. factory is delayed, by Bloomberg

BEFORE YOU GO

Lithium-ion battery danger. Researchers have found dangerous levels of PFAS “forever chemicals” near factories that make lithium-ion batteries, and Li-ion batteries are also causing major pollution in landfills, the Guardian reports. Li-ion batteries are integral to much of the clean energy transition and are therefore increasingly in use, so the researchers say it’s essential to step up environmental risk assessments around their production and disposal.