ChatGPT the lobbyist

Hey there, it’s Jillian in Brussels. OpenAI takes a swing at the political influence game. But first...Three things you need to know today:• [View in browser]( [Bloomberg]( Hey there, it’s Jillian in Brussels. OpenAI takes a swing at the political influence game. But first... Three things you need to know today: • US regulators [paused the Microsoft-Activision merger trial]( • Biden’s NSA nominee warned of [security threats from AI]( • Cyberattacks have [taken a toll on US colleges]( Hobby lobby The sudden rise of ChatGPT helped spur the European Union to quickly draw up rules around generative artificial intelligence. But the software’s creator, OpenAI, has been slow to influence the direction of the legislation. The EU is set to become the first Western government entity to implement mandatory AI regulations, with countries like the UK and [US trying to catch up](. The bloc’s three institutions met this week to negotiate a final deal, with the aim to reach an agreement by the end of the year. Google, IBM and Microsoft have been [lobbying EU officials from the get-go](, dating back to when the initial plan for the AI Act was proposed in April 2021. But OpenAI didn’t talk to a single European politician until June 2022. OpenAI decided to get involved when EU countries began pushing to add controls on general purpose AI, which would likely have ensnared the startup’s GPT technology. So OpenAI contacted a commission cabinet member, a couple of national government representatives and a pair of key lawmakers. ChatGPT didn’t yet exist at that point, and officials didn’t seem to take the company very seriously. One lawmaker canceled on OpenAI, and another didn’t even remember the meeting until staff triple-checked the agenda. The company, led by Sam Altman, seems to be having an easier time getting meetings today. Altman has met with top European politicians in Brussels and national capitals. But the world’s hottest tech startup is still learning how to engage with officials. This was on full display during a one-week span in May. Altman [urged a Senate panel to regulate AI](. Then he made headlines again a week later, when he said the company would [pull out of the EU]( if it couldn’t meet the bloc’s rules. Altman later explained: “I was like, ‘Of course we’re not going to break the law, so if we can’t comply, we won’t offer a service,’” he [told]( the Australian Financial Review. “I thought it was an innocuous answer, and it was twisted to be taken as a threat.” The moment should have taught Altman to be more cautious about what he says publicly ([just ask Mark Zuckerberg](). And it made EU politicians question OpenAI’s motivations. OpenAI appears to still be figuring out what it wants regulation to look like. Granted, it’s tricky. The EU’s three institutions still haven’t reached a consensus, and other tech companies have changed course at times. An OpenAI representative said the company has done a lot of policy research over the past year and has a clearer idea of what works. The company is getting some experts on the ground. It announced a new office in London, hired a European head of policy and partnerships in Germany and is looking to add someone in the UK. The new team has work to do. While the company has been [drawing up concrete commitments for the US government](, OpenAI has written only a handful of specific comments on the EU’s AI Act. Officials have told me that people at the company often contradict one another — all the way up to the chief executive officer. During a meeting late last month with Europe’s internal market commissioner, Thierry Breton, members of OpenAI’s technical team argued against the merits of watermarking, according to people present who asked not to be identified because they weren’t authorized to discuss the details publicly. Watermarking involves coding an AI to use certain words to make it easier to discern text generated by a computer from that of a human. OpenAI staff said at the meeting in San Francisco that watermarking is a limited way to moderate content. By the end of the day, Altman gave a public endorsement of watermarking. The CEO said on [video]( with the commissioner that he supports the practice. —[Jillian Deutsch](mailto:jdeutsch24@bloomberg.net) The big story Armando Pereira, the co-founder of telecom company Altice, [testified in a sweeping investigation]( into alleged corruption. He was detained in Portugal after police carried out 90 searches of homes and offices a week ago. One to watch [Watch the Bloomberg Technology TV interview]( with Cerebras CEO Andrew Feldman, discussing one of the world’s largest AI supercomputers. Get fully charged Taiwan Semiconductor Manufacturing cut its annual revenue outlook and [postponed the start of production]( at its signature Arizona project to 2025, twin setbacks for a chipmaking linchpin struggling with geopolitical tensions and a deep market slump. Google lost a bid to dismiss a lawsuit by an engineer who [claims he was fired for challenging a paper]( the company published that touted the ability of AI to speed the design of computer chips. The top US consumer financial watchdog is sharpening his focus on lenders’ use of AI when making credit decisions, signaling that [fresh restrictions loom]( for financial firms. A Chinese smartphone maker has its eye on [the Mexican market](. More from Bloomberg Get Bloomberg Tech newsletters in your inbox: - [Cyber Bulletin]( for coverage of the shadow world of hackers and cyber-espionage - [Game On]( for reporting on the video game business - [Power On]( for Apple scoops, consumer tech news and more - [Screentime]( for a front-row seat to the collision of Hollywood and Silicon Valley - [Soundbite]( for reporting on podcasting, the music industry and audio trends - [Hyperdrive]( for expert insight into the future of cars Follow Us Like getting this newsletter? [Subscribe to Bloomberg.com]( for unlimited access to trusted, data-driven journalism and subscriber-only insights. Want to sponsor this newsletter? [Get in touch here](. You received this message because you are subscribed to Bloomberg's Tech Daily newsletter. If a friend forwarded you this message, [sign up here]( to get it in your inbox. [Unsubscribe]( [Bloomberg.com]( [Contact Us]( Bloomberg L.P. 731 Lexington Avenue, New York, NY 10022 [Ads Powered By Liveintent]( [Ad Choices](