
Zuckerberg on compute per researcher being an underrated reason for researchers joining Meta's SuperIntelligence lab
Market Brief
Daily market recaps with key events, stock movements, and global influences
Meta invests $14.3B for 49% of Scale AI; weighs dropping open-source Behemoth AI; CEO Zuckerberg to testify in $8B Cambridge Analytica trial.
Removed 10M+ Facebook accounts for AI-driven spam and content duplication; stricter monetization rules for creatorsMeta Removes Over 10 Million Facebook Accounts in 2025 Content Strategy to Combat AI-Driven Duplication and Penalize Monetization
CEO Zuckerberg to testify in $8B shareholder trial over Cambridge Analytica privacy violationsMeta CEO Zuckerberg to Testify in $8 Billion Shareholder Trial Over Cambridge Analytica Privacy Violations
Considering shift from open-source Behemoth AI model to closed approachMeta Weighs Dropping Open-Source Behemoth AI for Closed Model
Threads app added direct messaging; daily active users near X’s, while X’s user base declinesMeta’s Threads Adds Direct Messaging, Narrows Gap With X
Developing proactive AI chatbots with Alignerr; targeting higher engagement and future AI revenueMeta Partners with Alignerr to Develop AI Studio Chatbots That Initiate and Remember User ConversationsMeta Tests 'Project Omni' Chatbots That Message Users First
Offering AI researchers up to $300M over four years to recruit top talent from rivalsMeta Offers Up to $300 Million Over Four Years, Over $100 Million First-Year Pay to Top AI Researchers Including OpenAI Staffers
Privacy concerns as Facebook AI can access private photos for personalized content (opt-in feature)Meta Enables Facebook AI to Access Private Photos With Cloud Processing for Collages, Recaps Amid Privacy Concerns
In talks to raise $29B in private credit for AI data centers; 2025 capex could reach $65BMeta Invests $14.3 Billion in Scale AI, Eyes $29 Billion Data-Center PushMeta Seeks $29 Billion Private Credit Deal for AI Data Centers
Facing EU and French antitrust scrutiny over ad practices and pay-or-consent model; risks daily finesFrance Antitrust Watchdog Alleges Meta Abused Power in Online AdsMeta Faces EU Scrutiny Over November 2024 Pay-or-Consent Model, Risks Daily Fines After €200 Million Digital Markets Act Fine, Plans Appeal
In advanced talks to acquire PlayAI, an AI voice startup, to enhance conversational AIMeta in Advanced Talks to Acquire Palo Alto-Based AI Voice Startup PlayAI, Following $14.3 Billion Scale AI Investment
Unveiled Oakley Meta HSTN AI glasses; acquired €3B stake in EssilorLuxottica for smart glasses expansionMeta Buys €3 B Stake in EssilorLuxottica to Expand Smart-Glasses PushMeta Unveils Oakley AI Glasses Starting at $399
Completed $14.3B investment for 49% of Scale AI; founder Wang joins as chief AI officerMeta Pays $14.3 Billion for 49% of Scale AI, Taps CEO to Run New LabMeta Pours $14.3 Billion Into Scale AI, Recruits Founder Alexandr Wang
Meta is intensifying its AI strategy, highlighted by the $14.3B investment in Scale AI and aggressive recruitment of top researchers, with compensation packages reaching $300M over four years. The company is also seeking $29B in private credit to fund a major buildout of AI data centers, signaling an unprecedented capital commitment to AI infrastructure. This push is designed to secure compute resources and talent as Meta consolidates its AI efforts under the new Superintelligence Labs, led by former Scale AI CEO Alexandr Wang.
Product development is moving quickly, with the launch of Oakley Meta HSTN smart glasses and a deeper stake in EssilorLuxottica, reinforcing Meta’s focus on AI-driven wearables. The company is also advancing AI chatbot capabilities through partnerships and internal projects, aiming to boost user engagement and lay groundwork for future AI monetization. Threads’ direct messaging rollout and rapid user growth are narrowing the gap with X, strengthening Meta’s social platform ecosystem.
However, the company faces significant regulatory and legal risks. EU and French authorities are scrutinizing Meta’s ad practices and pay-or-consent model, with the potential for daily fines. Privacy issues are in focus as Meta expands AI’s access to user content, and CEO Zuckerberg is set to testify in a high-profile $8B shareholder trial related to past privacy violations.
For traders, the near-term outlook hinges on Meta’s ability to execute its AI and hardware investments while managing regulatory headwinds. Watch for updates on the Scale AI fallout, progress on data center financing, regulatory developments in Europe, and signals from Q2 earnings on AI monetization and cost impacts. The stock’s muted reaction to recent news suggests investors are balancing optimism about AI growth with caution over spending and compliance risks.
Meta invests $14.3B for 49% of Scale AI; weighs dropping open-source Behemoth AI; CEO Zuckerberg to testify in $8B Cambridge Analytica trial.
Removed 10M+ Facebook accounts for AI-driven spam and content duplication; stricter monetization rules for creatorsMeta Removes Over 10 Million Facebook Accounts in 2025 Content Strategy to Combat AI-Driven Duplication and Penalize Monetization
CEO Zuckerberg to testify in $8B shareholder trial over Cambridge Analytica privacy violationsMeta CEO Zuckerberg to Testify in $8 Billion Shareholder Trial Over Cambridge Analytica Privacy Violations
Considering shift from open-source Behemoth AI model to closed approachMeta Weighs Dropping Open-Source Behemoth AI for Closed Model
Threads app added direct messaging; daily active users near X’s, while X’s user base declinesMeta’s Threads Adds Direct Messaging, Narrows Gap With X
Developing proactive AI chatbots with Alignerr; targeting higher engagement and future AI revenueMeta Partners with Alignerr to Develop AI Studio Chatbots That Initiate and Remember User ConversationsMeta Tests 'Project Omni' Chatbots That Message Users First
Offering AI researchers up to $300M over four years to recruit top talent from rivalsMeta Offers Up to $300 Million Over Four Years, Over $100 Million First-Year Pay to Top AI Researchers Including OpenAI Staffers
Privacy concerns as Facebook AI can access private photos for personalized content (opt-in feature)Meta Enables Facebook AI to Access Private Photos With Cloud Processing for Collages, Recaps Amid Privacy Concerns
In talks to raise $29B in private credit for AI data centers; 2025 capex could reach $65BMeta Invests $14.3 Billion in Scale AI, Eyes $29 Billion Data-Center PushMeta Seeks $29 Billion Private Credit Deal for AI Data Centers
Facing EU and French antitrust scrutiny over ad practices and pay-or-consent model; risks daily finesFrance Antitrust Watchdog Alleges Meta Abused Power in Online AdsMeta Faces EU Scrutiny Over November 2024 Pay-or-Consent Model, Risks Daily Fines After €200 Million Digital Markets Act Fine, Plans Appeal
In advanced talks to acquire PlayAI, an AI voice startup, to enhance conversational AIMeta in Advanced Talks to Acquire Palo Alto-Based AI Voice Startup PlayAI, Following $14.3 Billion Scale AI Investment
Unveiled Oakley Meta HSTN AI glasses; acquired €3B stake in EssilorLuxottica for smart glasses expansionMeta Buys €3 B Stake in EssilorLuxottica to Expand Smart-Glasses PushMeta Unveils Oakley AI Glasses Starting at $399
Completed $14.3B investment for 49% of Scale AI; founder Wang joins as chief AI officerMeta Pays $14.3 Billion for 49% of Scale AI, Taps CEO to Run New LabMeta Pours $14.3 Billion Into Scale AI, Recruits Founder Alexandr Wang
Meta is intensifying its AI strategy, highlighted by the $14.3B investment in Scale AI and aggressive recruitment of top researchers, with compensation packages reaching $300M over four years. The company is also seeking $29B in private credit to fund a major buildout of AI data centers, signaling an unprecedented capital commitment to AI infrastructure. This push is designed to secure compute resources and talent as Meta consolidates its AI efforts under the new Superintelligence Labs, led by former Scale AI CEO Alexandr Wang.
Product development is moving quickly, with the launch of Oakley Meta HSTN smart glasses and a deeper stake in EssilorLuxottica, reinforcing Meta’s focus on AI-driven wearables. The company is also advancing AI chatbot capabilities through partnerships and internal projects, aiming to boost user engagement and lay groundwork for future AI monetization. Threads’ direct messaging rollout and rapid user growth are narrowing the gap with X, strengthening Meta’s social platform ecosystem.
However, the company faces significant regulatory and legal risks. EU and French authorities are scrutinizing Meta’s ad practices and pay-or-consent model, with the potential for daily fines. Privacy issues are in focus as Meta expands AI’s access to user content, and CEO Zuckerberg is set to testify in a high-profile $8B shareholder trial related to past privacy violations.
For traders, the near-term outlook hinges on Meta’s ability to execute its AI and hardware investments while managing regulatory headwinds. Watch for updates on the Scale AI fallout, progress on data center financing, regulatory developments in Europe, and signals from Q2 earnings on AI monetization and cost impacts. The stock’s muted reaction to recent news suggests investors are balancing optimism about AI growth with caution over spending and compliance risks.
7 posts • GPT (4.1 mini)
Published
Meta CEO Mark Zuckerberg outlined the company's vision for artificial intelligence, emphasizing the development of "personal superintelligence" to empower individuals globally. Zuckerberg highlighted that Meta's AI lab aims to deliver AI capabilities that provide users with a cognitive advantage, such as AI glasses that assist with follow-ups and memory retention. He noted that top AI researchers from competitors like Anthropic and OpenAI are attracted to Meta's AI unit due to its unlimited computing power and smaller team environments, rather than solely large compensation packages. Zuckerberg also mentioned that the availability of extensive compute resources per researcher is a key factor in attracting talent. Meta's approach focuses on enabling personal AI tools rather than competing on the scale of financial incentives. Recruiting strategies have shifted to attracting teams of AI researchers simultaneously, reflecting a more collaborative approach within the industry.
Zuckerberg on compute per researcher being an underrated reason for researchers joining Meta's SuperIntelligence lab
One theme is loud and clear from 1) Zuckerberg describing AI recruiting to me 2) the windsurf deal and 3) the recent Meta raid on OpenAI Recruiting is now 1:many. AI researchers come as a team.
Mark Zuckerberg Q&A: Meta's focus on “personal super intelligence”, how large compensation packages pale in comparison to AI infra investments, AI glasses, more (@jessicalessin/The Information)
Meta CEO Mark Zuckerberg claims top talent from Anthropic and OpenAI are excited to join the company's new AI unit due to infinite access to computing power and small teams (not just HUGE signing bonuses) 😏
"Our mission with the lab is to deliver personal superintelligence to everyone in the world. So that way, we can put that power in every individual's hand." - Mark Watch Mark's full interview with The Information as he goes deeper on Meta's vision for superintelligence and
8 posts • GPT (4.1 mini)
Published
Meta Platforms Inc. has undertaken its largest content cleanup of the year by removing over 10 million Facebook profiles to address issues related to spam, impersonation, and AI-driven content duplication. The company announced new measures targeting accounts that repeatedly reuse photos, videos, or text from other users, labeling such posts as "unoriginal." Under the updated content strategy for 2025, Facebook creators who steal and repost videos risk losing monetization privileges. Meta's crackdown also targets accounts that share auto-generated captions, low-quality AI videos, or apply their own watermarks to others' content. These steps are part of a broader effort by Meta and other platforms like YouTube to curb the spread of non-original content and promote authentic creation on social media.
5 posts • OpenAI (o3)
Published
The Gauteng High Court in Johannesburg has ordered Meta Platforms to permanently shut down more than 30 anonymous Instagram handles and WhatsApp channels that have been distributing explicit child pornography depicting South African schoolchildren. Judge Mudunwazi Makamu gave the company until midday on Tuesday, 15 July, to disable the accounts and to hand over all information identifying those who created and operate them.
Read more
While the Gauteng High Court, Johannesburg, on Monday evening ordered Meta to shut down several anonymous Instagram accounts and WhatsApp channels, which are distributing explicit child pornography depicting South African school children, this has not yet been done by early
Meta Platforms has been given until midday on Tuesday to permanently remove more than 30 Instagram handles and WhatsApp channel accounts posting graphic pornography involving schoolchildren and provide information about who is behind them.
The Gauteng High Court in Johannesburg has ordered WhatsApp and Instagram parent company Meta to permanently shut down channels publishing sexual content involving South African schoolchildren and identify the person behind them. | @karynmaughan
The Gauteng High Court in Johannesburg has ordered tech giant Meta to shut down certain anonymous Instagram accounts and WhatsApp Channels distributing explicit child pornography involving South African schoolchildren.
Come up with simple mechanism to remove private images and videos of women from internet, Madras High Court directs Centre, @imranhindu reports.
28 posts • OpenAI (o3)
Published
Meta Platforms Chief Executive Officer Mark Zuckerberg has outlined the company’s most ambitious infrastructure build-out to date, pledging to spend about $70 billion this year on artificial-intelligence hardware and facilities and, over time, “hundreds of billions of dollars” to pursue what he called superintelligence.
The programme centres on a series of multi-gigawatt data-centre clusters. Prometheus, a one-gigawatt facility in New Albany, Ohio, is scheduled to begin operating in 2026. A second complex, dubbed Hyperion, will be erected in Louisiana, initially delivering two gigawatts of capacity by 2030 and scaling to as much as five gigawatts—roughly the power demand of a small city and large enough to span most of Manhattan.
Read more
Zuckerberg on compute per researcher being an underrated reason for researchers joining Meta's SuperIntelligence lab
Mark Zuckerberg Q&A: Meta's focus on “personal super intelligence”, how large compensation packages pale in comparison to AI infra investments, AI glasses, more (@jessicalessin/The Information)
Could AI glasses be your next superpower? Meta CEO Mark Zuckerberg envisions a future where these smart companions give you a "cognitive advantage," handling follow-ups and remembering details. The next evolution of personal tech. Watch the full episode on The Information's
Meta's Hyperion Data Center: "Meta Superintelligence Labs will have industry-leading levels of compute"
⚡ Meta’s building a 5GW AI supercenter called Hyperion. Yes, gigawatt. The AI arms race just went nuclear.
19 posts • OpenAI (o3)
Published
The European Commission has issued detailed guidelines under Article 28 of the Digital Services Act that allow EU member states to set their own minimum age for access to social-media platforms. The document confirms that outright bans for users below a chosen threshold—15 years in France’s case—would be considered compatible with EU law, provided they are backed by effective age-verification measures.
Read more
5 posts • GPT (4.1 mini)
Published
PhonePe, India's largest digital payments platform, has appointed Shivnath Thukral, former Public Policy Director at Meta India, as Vice President of Public Policy and Government Affairs. In his new role, Thukral will join PhonePe's leadership team and work closely with the company's founders. This strategic hire comes as PhonePe prepares for its initial public offering (IPO).
13 posts • GPT (4.1 mini)
Published
Meta Platforms Inc. CEO Mark Zuckerberg is set to appear as the star witness in an $8 billion trial beginning this week, brought by shareholders alleging privacy violations related to the 2018 Cambridge Analytica scandal. The lawsuit accuses Zuckerberg of operating Facebook as an illegal enterprise that allowed unauthorized harvesting of users' data, breaching promises to protect user information while profiting from the situation. Former COO Sheryl Sandberg and other board members, including Marc Andreessen and Peter Thiel, are also expected to testify. The case represents one of the largest privacy-related shareholder claims against a public company and raises questions about board oversight failures during the scandal. The trial is being closely watched as it may pose risks for Delaware’s business court system, which is handling the case.
8 posts • GPT (4.1 mini)
Published
The U.S. Department of Justice's Federal Programs Branch, responsible for defending former President Donald Trump's administration policies in court, has experienced a substantial reduction in staff, with nearly two-thirds of its lawyers resigning since the 2024 election. Former staff members have described the work as demoralizing, citing difficulties in defending policies they considered indefensible and contradictory to their expectations of supporting effective government functioning. Concurrently, Meta Platforms Inc.
and its CEO Mark Zuckerberg are preparing to face an $8 billion trial brought by investors alleging privacy violations. The lawsuit accuses Zuckerberg of managing Facebook in a way that compromised user data protections and profited from the resulting issues. The case represents a significant legal challenge for Meta, with shareholders seeking substantial financial remedies.
9 posts • OpenAI (o3)
Published
Meta Platforms’ newly formed Superintelligence Lab is debating whether to abandon Behemoth, its most powerful open-source artificial-intelligence model, in favor of developing a closed system, according to people briefed on internal discussions cited by the New York Times. The talks, led by 28-year-old Chief AI Officer Alexandr Wang, follow the company’s decision last month to consolidate its most advanced AI work inside the lab.
A move away from open source would mark a sharp philosophical shift for Meta, which has long argued that publicly released code accelerates industrywide innovation. The deliberations are still preliminary and would require approval from Chief Executive Officer Mark Zuckerberg. Meta has reportedly finished training Behemoth but delayed its release after weak internal test results, as the company races to keep pace with rivals such as Google, OpenAI and Anthropic.
$META's new Superintelligence Lab, led by Chief AI Officer Alexandr Wang, has discussed dropping its most advanced open-source model, Behemoth, in favor of a closed approach, according to The New York Times.
New York Time @nytimes: Meta's new Superintelligence Lab, led by Alexandr Wang, is considering a shift from open-source A.I. to a closed model. This realignment may impact global AI innovation, including within the competitive Nordic and European ecosys…
$META ’S NEW CHIEF AI OFFICER IS CONSIDERING SCRAPPING ITS MOST POWERFUL OPEN-SOURCE AI MODEL, BEHEMOTH IN FAVOR OF CLOSED MODEL - NYT
META’S NEW SUPERINTELLIGENCE LAB WEIGHING STRATEGIC SHIFT, MAY ABANDON OPEN-SOURCE AI MODEL ‘BEHEMOTH’ UNDER NEW CHIEF AI OFFICER — NYT $META
META’S NEW CHIEF AI OFFICER IS CONSIDERING SCRAPPING ITS MOST POWERFUL OPEN-SOURCE AI MODEL, BEHEMOTH — NYT.
5 posts • OpenAI (o3)
Published
ByteDance has begun developing a lightweight mixed-reality headset that resembles a pair of goggles and is tethered to a separate compute puck, according to people familiar with the matter cited by The Information. The company has formed an internal team to design custom chips for the device in an effort to improve performance and reduce power consumption.
The project positions the TikTok owner to compete more directly with Meta Platforms, which is working on a comparable slimline headset expected to arrive in 2027. A successful launch would broaden ByteDance’s hardware ambitions beyond its Pico virtual-reality unit, acquired in 2021, and intensify a growing race among large consumer-tech firms to make head-worn devices lighter, cheaper and more appealing to mass-market users.
ByteDance has not disclosed a timetable for commercial release or pricing, and the specifications could still change as development progresses. Nonetheless, the move underscores the company’s commitment to establish a footprint in augmented and mixed reality as the sector shifts toward thinner eyewear-style products.
Sources: ByteDance is working on a lightweight MR device that resembles goggles, tethered to a compute puck, and has a team working on the device's custom chips (The Information)
ByteDance is reportedly developing new, lightweight mixed reality goggles, challenging Meta's similar device. This move indicates a shift in the industry toward more consumer-friendly designs. Learn about the new direction in AR/VR: #mixedreality
Just in: ByteDance, TikTok's parent company, is developing lightweight mixed reality goggles, challenging Meta's $META similar product slated for 2027. #TechNews #Innovation
📢 𝐉𝐔𝐒𝐓 𝐈𝐍: Bytedance Developing Mixed Reality Goggles in Challenge to $META Meta- The Information
Have you read about XPANCEO, the company making smart contact lenses, which just got $250M of investment? Discover this and all the other top XR news of the week in my latest newsletter episode! #VirtualReality #AugmentedReality #technology
15 posts • GPT (4.1 mini)
Published
Meta Platforms is facing renewed scrutiny from the European Union over its pay-or-consent advertising model. Despite a €200 million ($232 million) fine imposed for breaching the EU's Digital Markets Act, Meta has not made sufficient changes to comply with the antitrust order issued in April. The European Commission has indicated that the social media giant's proposed adjustments to the model are limited and may not meet regulatory requirements. As a result, Meta could face daily fines if it fails to bring its pay-or-consent model into compliance in the coming weeks. Sources report that Meta is unlikely to propose further modifications to the model, increasing the risk of fresh antitrust charges and regulatory actions. The ongoing dispute highlights the EU's intensified efforts to enforce digital market regulations on major technology companies.
11 posts • OpenAI (o3)
Published
Meta Platforms is unlikely to offer any further concessions to the European Commission over its “pay-or-consent” advertising model, according to people familiar with the matter cited by Reuters on 11 July. The stance sets the owner of Facebook and Instagram on a collision course with EU regulators, who last month warned that only limited tweaks submitted by the company might not satisfy an April antitrust order issued under the Digital Markets Act.
The Commission fined Meta €200 million in April for operating the model—which lets users pay to avoid targeted ads or consent to data tracking—between November 2023 and November 2024. On 27 June the EU executive reminded the company that continued non-compliance could trigger daily penalties of up to 5 percent of Meta’s average worldwide turnover.
Sources now expect the Commission to file fresh charges within weeks, followed quickly by periodic fines if Meta does not revise its approach. Meta reiterated that it believes its service already exceeds DMA requirements and accused Brussels of discriminating against the company; the Commission declined to comment. Meta shares fell about 1.7 percent after the Reuters report.
Reuters reports $META is unlikely to make further changes to its pay-or-consent model, and the EU is preparing fresh charges under the Digital Markets Act. Daily fines could follow in the coming weeks if compliance isn’t met.
Meta won't tweak pay-or-consent model further despite risk of EU fines, sources say
📢 𝐉𝐔𝐒𝐓 𝐈𝐍: $META Meta Platforms Unlikely to Offer Additional Changes to Pay-or-Consent Model to EU - Reuters
Meta Is Not Expected To Make More Changes To Its Pay-Or-Consent Plan For The EU, But Could Face New EU Fines Under Digital Markets Act Due To Its Business Strategy, A Source Says 💻⚖️
European Commission says Meta has proposed limited changes to its pay-or-consent ad model, and Meta faces daily fines if it determines they are not sufficient (Reuters)
3 posts • GPT (4.1 mini)
Published
JPMorgan has raised its price targets for several major technology companies, reflecting improved channel checks and favorable currency movements amid lower recession risks. The bank increased Amazon's price target to $255 from $240 while maintaining an Overweight rating. Alphabet's price target was raised to $200 from $195, also with an Overweight rating. Netflix's price target was slightly increased to $1,230 from $1,220, with a Neutral rating.
Separately, Morgan Stanley raised its price target for Amazon to a street-high $300 from $250, maintaining an Overweight rating and naming it a top pick. The Morgan Stanley analyst cited a 9% and 6% increase in 2026 and 2027 EPS estimates, respectively, due to lower tariffs and growth drivers such as Anthropic and Azure's generative AI capabilities. Piper Sandler also raised Netflix's price target to $1,400 from $1,150, maintaining an Overweight rating, based on stronger revenue projections and improved margins expected from the third quarter of 2025 onward.
9 posts • GPT (4.1 mini)
Published
The Supreme Federal Court of Brazil (STF) has established what its president, Luiz Roberto Barroso, described as the world's most advanced regulation for social media platforms, setting parameters for the accountability of digital platforms. Minister Alexandre de Moraes emphasized the necessity of regulating social networks and holding major technology companies responsible, noting that freedom of expression has been misused to justify illicit activities. Moraes also linked big tech companies to the events of January 8, underscoring the STF's regulatory approach as an example for the world. This judicial stance has reportedly influenced former U.S.
President Donald Trump's decision to impose retaliatory measures against Brazil, with some U.S. allies viewing the STF ruling as contrary to freedom of expression. Additionally, a study by USP, UnB, and FGV revealed that Brazilian federal, state, and municipal governments spent R$9.3 billion in one year on products and services from big tech companies, challenging the notion that Brazil lacks resources to invest in technology. Privacy advocates, such as Cindy Cohn from the Electronic Frontier Foundation, warn that tech companies' cooperation with government requests is crucial to prevent excessive surveillance and repression.
11 posts • OpenAI (o3)
Published
Meta Platforms is escalating the race for artificial-intelligence talent, offering former Apple executive Ruoming Pang a compensation package exceeding $200 million to join its newly formed Meta Superintelligence Labs, according to multiple reports. Pang led Apple’s foundation-models group, a roughly 100-person team that developed large language models used across Siri and other Apple services.
The blockbuster deal is part of a broader hiring blitz directed by Chief Executive Officer Mark Zuckerberg. Meta has already recruited more than 10 researchers from OpenAI and specialists from Anthropic, Google and other rivals, with some candidates said to receive signing bonuses of about $100 million and total pay packages reaching $300 million over four years. Zuckerberg has also promised unrestricted access to advanced chips to entice researchers eager for computing power.
Read more
NEW: Mark Zuckerberg’s Hail Mary is paying a premium for some of the AI world's brightest minds to safeguard Meta’s future. But although its primary weapon is vast amounts of money, sources we talked to across the AI industry questioned whether it's enough.
Meta Offered Apple AI Executive Over $200 Million to Leave
Update from Bloomberg: Meta CEO Mark Zuckerberg has now successfully hired more than 10 OpenAI researchers, as well as top researchers and engineers from Anthropic, Google and other startups. Altman said that he hasn’t spoken to Zuckerberg since the Meta co-founder began his
Mark Zuckerberg is offering AI researchers vast amounts of compute — scientists need arrays of powerful chips to develop cutting-edge models — and with it the chance of being first to cross the superintelligence threshold, writes @parmy (via @opinion)
As the AI talent wars reach a fever pitch, Mark Zuckerberg is offering top tier recruits to Meta’s new superintelligence lab pay packages of up to $300 million over four years, with more than $100 million in total compensation for the first year.
5 posts • OpenAI (o3)
Published
France’s Competition Authority said on 9 July it has issued a formal statement of objections to Meta Platforms, indicating the US technology group may have breached national competition rules in the online advertising market.
Investigators allege Meta abused its dominant position by restricting advertisers’ access to third-party ad-verification partners—tools that help clients independently measure the performance and safety of their campaigns. The objections mark a preliminary step; Meta now has the opportunity to review the case file, submit written observations and request an oral hearing before any final decision is taken.
The regulator did not specify a timeline for the investigation’s next phase. Meta said it is reviewing the allegations and will cooperate with the authority’s proceedings.
4 posts • OpenAI (o3)
Published
The French government will finance a new coalition of civil-society organisations to flag hateful content directly to the national media regulator Arcom, Minister for the Fight Against Discrimination Aurore Bergé said on 9 July. The groups—including SOS Racisme, Licra, Le Crif, Osez le féminisme, Planning Familial and the Fédération des centres LGBTI+—will receive increased public funding to recruit staff dedicated to monitoring social-media platforms.
Under the plan, participating associations will have a priority channel to Arcom, enabling swifter removal of antisemitic, racist, anti-Muslim, homophobic and misogynist material posted online. Bergé described the initiative as a “democratic and public-health imperative” and warned digital platforms that France, in coordination with the European Union, is prepared to pursue sanctions if they do not meet their legal obligations.
40 posts • OpenAI (o3)
Published
Meta Platforms has acquired a stake of slightly below 3% in EssilorLuxottica for about €3 billion ($3.5 billion), according to people familiar with the matter cited by Bloomberg and Reuters. The transaction gives the U.S. social-media company an ownership position in the world’s largest eyewear maker, whose brands include Ray-Ban and Oakley.
The investment deepens a partnership that began in 2019 and has already produced two generations of Ray-Ban Meta smart glasses and, more recently, the sport-focused Oakley Meta HSTN model launched in June. Sources said Meta is considering raising its holding to roughly 5% over time as it seeks greater control over the design, manufacturing and distribution of AI-enabled wearable hardware.
Read more
13 posts • OpenAI (o3)
Published
OpenAI has hired four senior engineers from rival firms to strengthen its back-end infrastructure group, according to an internal Slack message from company co-founder Greg Brockman seen by WIRED. The recruits are David Lau, formerly Tesla’s vice president of software engineering; Uday Ruddarraju, former head of infrastructure engineering at xAI and social-media platform X; Mike Dalton, an infrastructure engineer who also worked at xAI; and Angela Fan, an artificial-intelligence researcher from Meta.
Read more
8 posts • GPT (4.1 mini)
Published
Meta has officially launched the Zoom app for its Meta Quest virtual reality headsets, making it available starting July 8, 2025. The standalone app enables users to join or host Zoom meetings entirely in VR, where participants appear as their Meta Avatars. This integration allows interaction with other attendees regardless of whether they join from desktop, mobile, or web platforms. The Zoom app is accessible as a free 2D Android application through the Meta Horizon Store, encouraging immersive video calls in virtual reality environments.
7 posts • GPT (4.1 mini)
Published
Meta CEO Mark Zuckerberg outlined the company's vision for artificial intelligence, emphasizing the development of "personal superintelligence" to empower individuals globally. Zuckerberg highlighted that Meta's AI lab aims to deliver AI capabilities that provide users with a cognitive advantage, such as AI glasses that assist with follow-ups and memory retention. He noted that top AI researchers from competitors like Anthropic and OpenAI are attracted to Meta's AI unit due to its unlimited computing power and smaller team environments, rather than solely large compensation packages. Zuckerberg also mentioned that the availability of extensive compute resources per researcher is a key factor in attracting talent. Meta's approach focuses on enabling personal AI tools rather than competing on the scale of financial incentives. Recruiting strategies have shifted to attracting teams of AI researchers simultaneously, reflecting a more collaborative approach within the industry.
Zuckerberg on compute per researcher being an underrated reason for researchers joining Meta's SuperIntelligence lab
One theme is loud and clear from 1) Zuckerberg describing AI recruiting to me 2) the windsurf deal and 3) the recent Meta raid on OpenAI Recruiting is now 1:many. AI researchers come as a team.
Mark Zuckerberg Q&A: Meta's focus on “personal super intelligence”, how large compensation packages pale in comparison to AI infra investments, AI glasses, more (@jessicalessin/The Information)
Meta CEO Mark Zuckerberg claims top talent from Anthropic and OpenAI are excited to join the company's new AI unit due to infinite access to computing power and small teams (not just HUGE signing bonuses) 😏
"Our mission with the lab is to deliver personal superintelligence to everyone in the world. So that way, we can put that power in every individual's hand." - Mark Watch Mark's full interview with The Information as he goes deeper on Meta's vision for superintelligence and
8 posts • GPT (4.1 mini)
Published
Meta Platforms Inc. has undertaken its largest content cleanup of the year by removing over 10 million Facebook profiles to address issues related to spam, impersonation, and AI-driven content duplication. The company announced new measures targeting accounts that repeatedly reuse photos, videos, or text from other users, labeling such posts as "unoriginal." Under the updated content strategy for 2025, Facebook creators who steal and repost videos risk losing monetization privileges. Meta's crackdown also targets accounts that share auto-generated captions, low-quality AI videos, or apply their own watermarks to others' content. These steps are part of a broader effort by Meta and other platforms like YouTube to curb the spread of non-original content and promote authentic creation on social media.
5 posts • OpenAI (o3)
Published
The Gauteng High Court in Johannesburg has ordered Meta Platforms to permanently shut down more than 30 anonymous Instagram handles and WhatsApp channels that have been distributing explicit child pornography depicting South African schoolchildren. Judge Mudunwazi Makamu gave the company until midday on Tuesday, 15 July, to disable the accounts and to hand over all information identifying those who created and operate them.
Read more
While the Gauteng High Court, Johannesburg, on Monday evening ordered Meta to shut down several anonymous Instagram accounts and WhatsApp channels, which are distributing explicit child pornography depicting South African school children, this has not yet been done by early
Meta Platforms has been given until midday on Tuesday to permanently remove more than 30 Instagram handles and WhatsApp channel accounts posting graphic pornography involving schoolchildren and provide information about who is behind them.
The Gauteng High Court in Johannesburg has ordered WhatsApp and Instagram parent company Meta to permanently shut down channels publishing sexual content involving South African schoolchildren and identify the person behind them. | @karynmaughan
The Gauteng High Court in Johannesburg has ordered tech giant Meta to shut down certain anonymous Instagram accounts and WhatsApp Channels distributing explicit child pornography involving South African schoolchildren.
Come up with simple mechanism to remove private images and videos of women from internet, Madras High Court directs Centre, @imranhindu reports.
28 posts • OpenAI (o3)
Published
Meta Platforms Chief Executive Officer Mark Zuckerberg has outlined the company’s most ambitious infrastructure build-out to date, pledging to spend about $70 billion this year on artificial-intelligence hardware and facilities and, over time, “hundreds of billions of dollars” to pursue what he called superintelligence.
The programme centres on a series of multi-gigawatt data-centre clusters. Prometheus, a one-gigawatt facility in New Albany, Ohio, is scheduled to begin operating in 2026. A second complex, dubbed Hyperion, will be erected in Louisiana, initially delivering two gigawatts of capacity by 2030 and scaling to as much as five gigawatts—roughly the power demand of a small city and large enough to span most of Manhattan.
Read more
Zuckerberg on compute per researcher being an underrated reason for researchers joining Meta's SuperIntelligence lab
Mark Zuckerberg Q&A: Meta's focus on “personal super intelligence”, how large compensation packages pale in comparison to AI infra investments, AI glasses, more (@jessicalessin/The Information)
Could AI glasses be your next superpower? Meta CEO Mark Zuckerberg envisions a future where these smart companions give you a "cognitive advantage," handling follow-ups and remembering details. The next evolution of personal tech. Watch the full episode on The Information's
Meta's Hyperion Data Center: "Meta Superintelligence Labs will have industry-leading levels of compute"
⚡ Meta’s building a 5GW AI supercenter called Hyperion. Yes, gigawatt. The AI arms race just went nuclear.
19 posts • OpenAI (o3)
Published
The European Commission has issued detailed guidelines under Article 28 of the Digital Services Act that allow EU member states to set their own minimum age for access to social-media platforms. The document confirms that outright bans for users below a chosen threshold—15 years in France’s case—would be considered compatible with EU law, provided they are backed by effective age-verification measures.
Read more
5 posts • GPT (4.1 mini)
Published
PhonePe, India's largest digital payments platform, has appointed Shivnath Thukral, former Public Policy Director at Meta India, as Vice President of Public Policy and Government Affairs. In his new role, Thukral will join PhonePe's leadership team and work closely with the company's founders. This strategic hire comes as PhonePe prepares for its initial public offering (IPO).
13 posts • GPT (4.1 mini)
Published
Meta Platforms Inc. CEO Mark Zuckerberg is set to appear as the star witness in an $8 billion trial beginning this week, brought by shareholders alleging privacy violations related to the 2018 Cambridge Analytica scandal. The lawsuit accuses Zuckerberg of operating Facebook as an illegal enterprise that allowed unauthorized harvesting of users' data, breaching promises to protect user information while profiting from the situation. Former COO Sheryl Sandberg and other board members, including Marc Andreessen and Peter Thiel, are also expected to testify. The case represents one of the largest privacy-related shareholder claims against a public company and raises questions about board oversight failures during the scandal. The trial is being closely watched as it may pose risks for Delaware’s business court system, which is handling the case.
8 posts • GPT (4.1 mini)
Published
The U.S. Department of Justice's Federal Programs Branch, responsible for defending former President Donald Trump's administration policies in court, has experienced a substantial reduction in staff, with nearly two-thirds of its lawyers resigning since the 2024 election. Former staff members have described the work as demoralizing, citing difficulties in defending policies they considered indefensible and contradictory to their expectations of supporting effective government functioning. Concurrently, Meta Platforms Inc.
and its CEO Mark Zuckerberg are preparing to face an $8 billion trial brought by investors alleging privacy violations. The lawsuit accuses Zuckerberg of managing Facebook in a way that compromised user data protections and profited from the resulting issues. The case represents a significant legal challenge for Meta, with shareholders seeking substantial financial remedies.
9 posts • OpenAI (o3)
Published
Meta Platforms’ newly formed Superintelligence Lab is debating whether to abandon Behemoth, its most powerful open-source artificial-intelligence model, in favor of developing a closed system, according to people briefed on internal discussions cited by the New York Times. The talks, led by 28-year-old Chief AI Officer Alexandr Wang, follow the company’s decision last month to consolidate its most advanced AI work inside the lab.
A move away from open source would mark a sharp philosophical shift for Meta, which has long argued that publicly released code accelerates industrywide innovation. The deliberations are still preliminary and would require approval from Chief Executive Officer Mark Zuckerberg. Meta has reportedly finished training Behemoth but delayed its release after weak internal test results, as the company races to keep pace with rivals such as Google, OpenAI and Anthropic.
$META's new Superintelligence Lab, led by Chief AI Officer Alexandr Wang, has discussed dropping its most advanced open-source model, Behemoth, in favor of a closed approach, according to The New York Times.
New York Time @nytimes: Meta's new Superintelligence Lab, led by Alexandr Wang, is considering a shift from open-source A.I. to a closed model. This realignment may impact global AI innovation, including within the competitive Nordic and European ecosys…
$META ’S NEW CHIEF AI OFFICER IS CONSIDERING SCRAPPING ITS MOST POWERFUL OPEN-SOURCE AI MODEL, BEHEMOTH IN FAVOR OF CLOSED MODEL - NYT
META’S NEW SUPERINTELLIGENCE LAB WEIGHING STRATEGIC SHIFT, MAY ABANDON OPEN-SOURCE AI MODEL ‘BEHEMOTH’ UNDER NEW CHIEF AI OFFICER — NYT $META
META’S NEW CHIEF AI OFFICER IS CONSIDERING SCRAPPING ITS MOST POWERFUL OPEN-SOURCE AI MODEL, BEHEMOTH — NYT.
5 posts • OpenAI (o3)
Published
ByteDance has begun developing a lightweight mixed-reality headset that resembles a pair of goggles and is tethered to a separate compute puck, according to people familiar with the matter cited by The Information. The company has formed an internal team to design custom chips for the device in an effort to improve performance and reduce power consumption.
The project positions the TikTok owner to compete more directly with Meta Platforms, which is working on a comparable slimline headset expected to arrive in 2027. A successful launch would broaden ByteDance’s hardware ambitions beyond its Pico virtual-reality unit, acquired in 2021, and intensify a growing race among large consumer-tech firms to make head-worn devices lighter, cheaper and more appealing to mass-market users.
ByteDance has not disclosed a timetable for commercial release or pricing, and the specifications could still change as development progresses. Nonetheless, the move underscores the company’s commitment to establish a footprint in augmented and mixed reality as the sector shifts toward thinner eyewear-style products.
Sources: ByteDance is working on a lightweight MR device that resembles goggles, tethered to a compute puck, and has a team working on the device's custom chips (The Information)
ByteDance is reportedly developing new, lightweight mixed reality goggles, challenging Meta's similar device. This move indicates a shift in the industry toward more consumer-friendly designs. Learn about the new direction in AR/VR: #mixedreality
Just in: ByteDance, TikTok's parent company, is developing lightweight mixed reality goggles, challenging Meta's $META similar product slated for 2027. #TechNews #Innovation
📢 𝐉𝐔𝐒𝐓 𝐈𝐍: Bytedance Developing Mixed Reality Goggles in Challenge to $META Meta- The Information
Have you read about XPANCEO, the company making smart contact lenses, which just got $250M of investment? Discover this and all the other top XR news of the week in my latest newsletter episode! #VirtualReality #AugmentedReality #technology
15 posts • GPT (4.1 mini)
Published
Meta Platforms is facing renewed scrutiny from the European Union over its pay-or-consent advertising model. Despite a €200 million ($232 million) fine imposed for breaching the EU's Digital Markets Act, Meta has not made sufficient changes to comply with the antitrust order issued in April. The European Commission has indicated that the social media giant's proposed adjustments to the model are limited and may not meet regulatory requirements. As a result, Meta could face daily fines if it fails to bring its pay-or-consent model into compliance in the coming weeks. Sources report that Meta is unlikely to propose further modifications to the model, increasing the risk of fresh antitrust charges and regulatory actions. The ongoing dispute highlights the EU's intensified efforts to enforce digital market regulations on major technology companies.
11 posts • OpenAI (o3)
Published
Meta Platforms is unlikely to offer any further concessions to the European Commission over its “pay-or-consent” advertising model, according to people familiar with the matter cited by Reuters on 11 July. The stance sets the owner of Facebook and Instagram on a collision course with EU regulators, who last month warned that only limited tweaks submitted by the company might not satisfy an April antitrust order issued under the Digital Markets Act.
The Commission fined Meta €200 million in April for operating the model—which lets users pay to avoid targeted ads or consent to data tracking—between November 2023 and November 2024. On 27 June the EU executive reminded the company that continued non-compliance could trigger daily penalties of up to 5 percent of Meta’s average worldwide turnover.
Sources now expect the Commission to file fresh charges within weeks, followed quickly by periodic fines if Meta does not revise its approach. Meta reiterated that it believes its service already exceeds DMA requirements and accused Brussels of discriminating against the company; the Commission declined to comment. Meta shares fell about 1.7 percent after the Reuters report.
Reuters reports $META is unlikely to make further changes to its pay-or-consent model, and the EU is preparing fresh charges under the Digital Markets Act. Daily fines could follow in the coming weeks if compliance isn’t met.
Meta won't tweak pay-or-consent model further despite risk of EU fines, sources say
📢 𝐉𝐔𝐒𝐓 𝐈𝐍: $META Meta Platforms Unlikely to Offer Additional Changes to Pay-or-Consent Model to EU - Reuters
Meta Is Not Expected To Make More Changes To Its Pay-Or-Consent Plan For The EU, But Could Face New EU Fines Under Digital Markets Act Due To Its Business Strategy, A Source Says 💻⚖️
European Commission says Meta has proposed limited changes to its pay-or-consent ad model, and Meta faces daily fines if it determines they are not sufficient (Reuters)
3 posts • GPT (4.1 mini)
Published
JPMorgan has raised its price targets for several major technology companies, reflecting improved channel checks and favorable currency movements amid lower recession risks. The bank increased Amazon's price target to $255 from $240 while maintaining an Overweight rating. Alphabet's price target was raised to $200 from $195, also with an Overweight rating. Netflix's price target was slightly increased to $1,230 from $1,220, with a Neutral rating.
Separately, Morgan Stanley raised its price target for Amazon to a street-high $300 from $250, maintaining an Overweight rating and naming it a top pick. The Morgan Stanley analyst cited a 9% and 6% increase in 2026 and 2027 EPS estimates, respectively, due to lower tariffs and growth drivers such as Anthropic and Azure's generative AI capabilities. Piper Sandler also raised Netflix's price target to $1,400 from $1,150, maintaining an Overweight rating, based on stronger revenue projections and improved margins expected from the third quarter of 2025 onward.
9 posts • GPT (4.1 mini)
Published
The Supreme Federal Court of Brazil (STF) has established what its president, Luiz Roberto Barroso, described as the world's most advanced regulation for social media platforms, setting parameters for the accountability of digital platforms. Minister Alexandre de Moraes emphasized the necessity of regulating social networks and holding major technology companies responsible, noting that freedom of expression has been misused to justify illicit activities. Moraes also linked big tech companies to the events of January 8, underscoring the STF's regulatory approach as an example for the world. This judicial stance has reportedly influenced former U.S.
President Donald Trump's decision to impose retaliatory measures against Brazil, with some U.S. allies viewing the STF ruling as contrary to freedom of expression. Additionally, a study by USP, UnB, and FGV revealed that Brazilian federal, state, and municipal governments spent R$9.3 billion in one year on products and services from big tech companies, challenging the notion that Brazil lacks resources to invest in technology. Privacy advocates, such as Cindy Cohn from the Electronic Frontier Foundation, warn that tech companies' cooperation with government requests is crucial to prevent excessive surveillance and repression.
11 posts • OpenAI (o3)
Published
Meta Platforms is escalating the race for artificial-intelligence talent, offering former Apple executive Ruoming Pang a compensation package exceeding $200 million to join its newly formed Meta Superintelligence Labs, according to multiple reports. Pang led Apple’s foundation-models group, a roughly 100-person team that developed large language models used across Siri and other Apple services.
The blockbuster deal is part of a broader hiring blitz directed by Chief Executive Officer Mark Zuckerberg. Meta has already recruited more than 10 researchers from OpenAI and specialists from Anthropic, Google and other rivals, with some candidates said to receive signing bonuses of about $100 million and total pay packages reaching $300 million over four years. Zuckerberg has also promised unrestricted access to advanced chips to entice researchers eager for computing power.
Read more
NEW: Mark Zuckerberg’s Hail Mary is paying a premium for some of the AI world's brightest minds to safeguard Meta’s future. But although its primary weapon is vast amounts of money, sources we talked to across the AI industry questioned whether it's enough.
Meta Offered Apple AI Executive Over $200 Million to Leave
Update from Bloomberg: Meta CEO Mark Zuckerberg has now successfully hired more than 10 OpenAI researchers, as well as top researchers and engineers from Anthropic, Google and other startups. Altman said that he hasn’t spoken to Zuckerberg since the Meta co-founder began his
Mark Zuckerberg is offering AI researchers vast amounts of compute — scientists need arrays of powerful chips to develop cutting-edge models — and with it the chance of being first to cross the superintelligence threshold, writes @parmy (via @opinion)
As the AI talent wars reach a fever pitch, Mark Zuckerberg is offering top tier recruits to Meta’s new superintelligence lab pay packages of up to $300 million over four years, with more than $100 million in total compensation for the first year.
5 posts • OpenAI (o3)
Published
France’s Competition Authority said on 9 July it has issued a formal statement of objections to Meta Platforms, indicating the US technology group may have breached national competition rules in the online advertising market.
Investigators allege Meta abused its dominant position by restricting advertisers’ access to third-party ad-verification partners—tools that help clients independently measure the performance and safety of their campaigns. The objections mark a preliminary step; Meta now has the opportunity to review the case file, submit written observations and request an oral hearing before any final decision is taken.
The regulator did not specify a timeline for the investigation’s next phase. Meta said it is reviewing the allegations and will cooperate with the authority’s proceedings.
4 posts • OpenAI (o3)
Published
The French government will finance a new coalition of civil-society organisations to flag hateful content directly to the national media regulator Arcom, Minister for the Fight Against Discrimination Aurore Bergé said on 9 July. The groups—including SOS Racisme, Licra, Le Crif, Osez le féminisme, Planning Familial and the Fédération des centres LGBTI+—will receive increased public funding to recruit staff dedicated to monitoring social-media platforms.
Under the plan, participating associations will have a priority channel to Arcom, enabling swifter removal of antisemitic, racist, anti-Muslim, homophobic and misogynist material posted online. Bergé described the initiative as a “democratic and public-health imperative” and warned digital platforms that France, in coordination with the European Union, is prepared to pursue sanctions if they do not meet their legal obligations.
40 posts • OpenAI (o3)
Published
Meta Platforms has acquired a stake of slightly below 3% in EssilorLuxottica for about €3 billion ($3.5 billion), according to people familiar with the matter cited by Bloomberg and Reuters. The transaction gives the U.S. social-media company an ownership position in the world’s largest eyewear maker, whose brands include Ray-Ban and Oakley.
The investment deepens a partnership that began in 2019 and has already produced two generations of Ray-Ban Meta smart glasses and, more recently, the sport-focused Oakley Meta HSTN model launched in June. Sources said Meta is considering raising its holding to roughly 5% over time as it seeks greater control over the design, manufacturing and distribution of AI-enabled wearable hardware.
Read more
13 posts • OpenAI (o3)
Published
OpenAI has hired four senior engineers from rival firms to strengthen its back-end infrastructure group, according to an internal Slack message from company co-founder Greg Brockman seen by WIRED. The recruits are David Lau, formerly Tesla’s vice president of software engineering; Uday Ruddarraju, former head of infrastructure engineering at xAI and social-media platform X; Mike Dalton, an infrastructure engineer who also worked at xAI; and Angela Fan, an artificial-intelligence researcher from Meta.
Read more
8 posts • GPT (4.1 mini)
Published
Meta has officially launched the Zoom app for its Meta Quest virtual reality headsets, making it available starting July 8, 2025. The standalone app enables users to join or host Zoom meetings entirely in VR, where participants appear as their Meta Avatars. This integration allows interaction with other attendees regardless of whether they join from desktop, mobile, or web platforms. The Zoom app is accessible as a free 2D Android application through the Meta Horizon Store, encouraging immersive video calls in virtual reality environments.