Welcome to The Tech Fix, a weekly newsletter where we break down the latest technology news and trends from the advertising and marketing industry, curated by technology editor Jessica Heygate.

A group of AI scientists, professors and tech leaders have banded together to issue a warning about the catastrophic risks of AI technology.

More than 300 people have signed a statement published by the U.S.-based nonprofit the Center for AI Safety on Tuesday that aims to succinctly surmise concerns about AI and “open up discussion.”

“Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war,” the statement reads.

Signatories include OpenAI CEO Sam Altman, Google DeepMind CEO Demis Hassabis, Anthropic CEO Dario Amodei and Microsoft CTO Kevin Scott. Meta, which recently rolled out a host of generative AI tools for advertising, was noticeably absent from the list.

OpenAI last week announced it would award ten $100,000 grants to fund experiments in democratic AI governance processes. The developer of popular AI chatbot ChatGPT has continuously pushed for AI regulation to “mitigate the risks of increasingly powerful models,” Altman said before a Senate committee hearing to discuss oversight of AI earlier this month. In a blog post last week, three OpenAI cofounders called for an international regulatory agency similar to the IAEA, which governs nuclear power.

OpenAI investor Microsoft has also proposed a range of regulations for AI, including a requirement for systems used in critical infrastructure to have a kill switch.

The regulatory calls come amid a flurry of new generative AI tools, including Google’s conversational ad platform, and the continued explosive growth of ChatGPT following the release of its iOS app. The app topped U.S. download charts for three days and surpassed half a million downloads in its first six days, according to Data.ai. 

• EXCLUSIVE: From ad blocking to ad serving: Adblock Plus creator Eyeo is launching an ad solution for marketers to reach its database of 250 million monthly active users. Eyeo said the “majority” of its ad blocking users have consented to seeing non-intrusive online ads that meet its Acceptable Ads Standard. Based on this standard, Eyeo has built a marketplace of 3,500 publishers whose inventory will now be available to purchase through Eyeo’s programmatic ad solution via private marketplace deals.

• Twitter is making its ad inventory available to purchase on the open programmatic marketplace for the first time via a partnership with InMobi, Digiday reports. It's also extending Community Notes, its crowdsourced content moderation mechanism, to images, a move it hopes will help it tackle the proliferation of “misleading media” on the platform including AI-generated and manipulated content. The ‘Notes on Media’ tool is currently in pilot and will be limited to its top-rated contributors.

• Twitter also launched a new API tier for startups that costs $5,000 per month — but some users have argued the tier is too expensive and comes too late, after several developers sold or shuttered their businesses when Musk cut free API access.

• Instacart has rolled out an AI-powered search tool, Ask Instacart, powered by ChatGPT. First announced in March during OpenAI’s ChatGPT API announcement, it allows Instacart shoppers to ask questions in natural language and receive extended results beyond product recommendations, including information about food preparation, product attributes and dietary considerations.

• TikTok is testing an AI chatbot called Tako that can surface information about a video or provide recommendations.

• Snapchat has added new capability to My AI that allows Snapchat+ subscribers to send images to the chatbot, which it will respond to. For instance, a user could share a picture of their grocery haul and the AI may respond with a recipe recommendation.

• Instagram has officially launched ads in search results via the Instagram Marketing API two months after it said it was testing the new format.

• YouTube is shutting down its Stories format on June 26 in a move to prioritize Shorts, Community posts, live video and long-form video, the company announced last week.

• Oracle has terminated social bookmarking service AddThis, which it acquired for $200 million in 2016 to augment its Data Cloud business. Oracle has been reducing its use of AddThis data since 2019 over privacy concerns, Adexchanger reports.

WPP teamed up with Nvidia to build a generative AI-powered content engine that it claims will produce immersive, tailored and more efficient content for clients.

European Commission vice president Margrethe Vestager said on Wednesday that Europe and the U.S. are drawing up a voluntary code of conduct for AI, with a draft expected within weeks.

Twitter has withdrawn from the European Union’s voluntary Code of Practice, which aims to eradicate online disinformation. The Code of Practice has 38 signatories, including major tech firms Google, Meta and TikTok.

European commissioner Thierry Breton confirmed that Twitter had pulled out of the agreement last week, alongside a warning to the platform.

“Twitter leaves EU voluntary Code of Practice against disinformation. But obligations remain. You can run but you can’t hide,” Breton wrote. He went on to reference that Twitter will be legally required to mitigate disinformation under the EU’s Digital Services Act (DSA), which carries extra requirements for so-called “very large online platforms.”

Minnesota governor Tim Walz vetoed a bill to guarantee a minimum wage and other protections for Uber and Lyft drivers. Walz said in a letter to the speaker of the Minnesota House of Representatives that “ride-share drivers deserve safe working conditions and fair wages,” but added that the proposed legislation “is not the right bill to achieve these goals.” California, New York City and Washington have all passed laws guaranteeing minimum wages for gig drivers.

DoorDash has responded to a lawsuit alleging the company deliberately charges iPhone users more than Android users, deeming the claims as “blatantly false” and “frankly ridiculous” in a fiery blog post.

The class-action lawsuit against DoorDash, filed in the U.S. District Court of Maryland, alleges the food delivery platform engages in “deceptive, misleading and fraudulent practices.”

• Meta has sold Giphy to Shutterstock for $53 million after the social media firm was ordered to discard its investment by the U.K.’s antitrust authority in October. The sale represents a huge markdown from the $400 million Meta paid for the GIF search engine in 2020. Shutterstock said it expects the deal, which includes an agreement to allow Meta to continue accessing Giphy’s content across its platforms, to close in June.

• Snowflake acquired search startup Neeva for an undisclosed sum. The deal came days after Neeva announced it was shutting down its consumer search product and shifting to focus on enterprise use cases of LLMs and generative AI.

Take a break from AI and be sure to read the surgeon general’s 19-page advisory about the risks of social media for young people. It isn’t all doom and gloom; it also includes practical tips for parents and caregivers to protect their children from harm, as well as recommendations for what policymakers, tech firms and researchers should do to help.