The Washington PostDemocracy Dies in Darkness
The Technology 202

A newsletter briefing on the intersection of technology and politics.

OpenAI embraced regulation— until talks got serious in Europe

The Technology 202

A newsletter briefing on the intersection of technology and politics.

Happy Friday! This intro was written before last night’s Heat-Celtics game, so send news tips and either congrats or condolences to: cristiano.lima@washpost.com.

Below: Twitter is likely to leave an E.U. disinformation pact, and Canada launches a privacy probe into OpenAI. First:

OpenAI embraced regulation — until talks got serious in Europe

Senators fawned over testimony from OpenAI CEO Sam Altman earlier this month that embraced regulation of artificial intelligence tools, like his popular ChatGPT chatbot. 

But days later, as European policymakers forge ahead with what could soon become the world’s most sweeping and significant AI rule book, the company is balking at the fine print. Altman this week threatened to pull their services from the E.U. over the proposed rules, a move that comes as other companies like Google tread lightly by declining to offer similar products in the region.

The shift may foreshadow a conflict stateside.

While lawmakers in the United States have expressed optimism about finding common ground with industry on AI regulation, they are likely barreling toward their own high-stakes confrontation over what it should look like. 

Altman floated several ideas for regulating AI at his first congressional hearing last Tuesday, including standing up a new regulator to license large AI models, creating safety requirements before products are rolled out and establishing a fresh liability regime for AI harms. 

“If this technology goes wrong, it can go quite wrong and we want to be vocal about that,” Altman said. “We want to work with the government to prevent that from happening.”

Senate Judiciary Chairman Richard J. Durbin (D-Ill.) said it was “historic” to have “people representing large corporations … come before us and plead with us to regulate them.”

Sen. Richard Blumenthal (D-Conn.), whose subcommittee hosted the session, said Altman’s commitment to “specific action” was “night and day compared to other CEOs.” 

But it’s a lot easier to find agreement when discussing regulatory principles than when hammering out legislative text and doling out legal obligations.

Officials overseas learned that the hard way this week as Altman threatened to pull OpenAI’s services from the European Union over the bloc’s proposed AI Act, which would create new transparency and safety requirements for AI developers. 

“The details really matter,” Altman told reporters during a trip to London this week, according to the Financial Times. “We will try to comply, but if we can’t comply we will cease operating.”

European officials recently proposed an expanded version of the measure that would put greater restrictions on general-purpose AI products like ChatGPT, including requiring that they disclose when they have used copyrighted data to train their models. 

That could open the door for content creators to seek compensation for their material, a general aim that Altman appeared to express support for during his recent Senate hearing. 

European leaders rebuked Altman’s remarks, with one top official accusing him of attempting to “blackmail” the bloc to thwart their regulatory efforts. 

“Let’s be clear, our rules are put in place for the security and well-being of our citizens and this cannot be bargained,” Thierry Breton, E.U. commissioner for the internal market, told Reuters.

Timnit Gebru, a prominent AI ethicist and former Google staffer, disputed Altman’s assertion the company would not be able to comply with the bloc’s rules. 

OpenAI isn’t the only prominent tech company scrambling to figure out its stance on Europe’s proposed rules. Google earlier this month rolled out an updated version of its AI chatbot Bard, a ChatGPT competitor, but notably did not make it available in the E.U., according to reports.

This week after meeting with Google CEO Sundar Pichai, European officials announced they plan to develop a voluntary AI framework that companies can follow before any rules hit the books, which could create opportunities for the two sides to hash out compliance issues.

The advanced discussions around AI regulation in Europe could foreshadow similar clashes in the United States as policymakers look to turn their legislative ideas into actionable proposals. 

Our top tabs

Twitter will likely stop participating in E.U. disinformation code of practice, official says

An E.U. official on Thursday said Twitter is likely to cease participating in a code of practice that helps the bloc tackle online disinformation, though the move does not signal it will stop operating in Europe, Foo Yun Chee reports for Reuters.

The guidelines, titled the E.U. Code of Practice on Disinformation, were first adopted in 2018 as a means of motivating companies to weed out disinformation on their platforms. Major tech companies, including TikTok and Meta, are signatories on the code.

The Reuters report adds: “Twitter has given signs that it will leave the code, the EU official said, adding that it does not make a big difference as the company has not been putting in a lot of effort recently.”

“It just means that they won't attend meetings and not issue reports. They would still have legal obligations,” said the official, who later added: “They are not pulling out of Europe.” 

Zuckerberg hopes for more stability at Meta following rounds of layoffs

Meta chief Mark Zuckerberg in a companywide meeting on Thursday told employees he hopes the company will have more stability with less bureaucracy in the future, our colleague Naomi Nix reports, citing a recording of the call.

“Going through restructuring and layoffs and changes like this is obviously a very difficult thing,” Zuckerberg said. “So it’s not like we’re going to end up in exactly the place that we were before because that wasn’t my goal. I wanted to get to a scrappier place.”

Some 5,100 people were laid off from Meta in May, according to human resources chief Lori Goler.

The company in March laid off recruiters followed by around 4,000 others in technical teams in late April. “The latest job cuts add to the November workforce reduction, which gutted about 11,000 jobs, or about 13 percent of Meta’s workforce,” Naomi writes.

Zuckerberg has previously said 2023 would be a “year of efficiency” for Meta as it faces tense competition in the ad revenue space and seeks to build out its Metaverse vision.

Canada probing ChatGPT-owned OpenAI over privacy concerns

Canadian regulators will launch an investigation into ChatGPT-owned OpenAI over data collection and usage concerns, Tiyashi Datta reports for Reuters.

Datta writes: “The federal privacy regulator, along with counterparts in Quebec, British Columbia and Alberta, will investigate if OpenAI has obtained consent for the collection, use and disclosure of personal information of residents via ChatGPT, the Office of the Privacy Commissioner of Canada said on Thursday.”

OpenAI did not immediately respond to a request for comment from Reuters.

Britain’s antitrust enforcer earlier this month opened a similar investigation into large-language and generative AI models to examine competition and consumer risk concerns in the AI space.

Global policymakers are making sweeping attempts at AI regulations amid a boom in popularity for AI tools like ChatGPT, Google’s Bard and Microsoft’s Bing AI chatbot. Altman, OpenAI’s CEO, testified before a Senate panel about the future of AI last week, while Microsoft’s Brad Smith on Thursday delivered an AI regulation pitch to an audience of U.S. lawmakers.

Inside the industry

How generative AI could generate more antisemitism (Axios)

TikTok is also making an AI chatbot, for some reason (Shira Ovide)

Dish in talks to sell wireless plans through Amazon (Wall Street Journal)

Meta faces pressure to support independent audit of risk oversight committee (The Hill)

Competition watch

Microsoft UK veto versus EU nod poses questions, Vestager says (Bloomberg News)

US push for controls on investment in China hits EU resistance (Bloomberg News)

Workforce report

A hiring law blazes a path for A.I. regulation (New York Times)

Trending

Why Nvidia is suddenly one of the most valuable companies in the world (Gerrit De Vynck)

Daybook

Before you log off

Thats all for today — thank you so much for joining us! Make sure to tell others to subscribe to The Technology 202 here. Get in touch with tips, feedback or greetings on Twitter or email

Loading...