Yazılar

Microsoft May Walk Away from OpenAI Negotiations Amid Stake Disputes

Microsoft is reportedly prepared to abandon high-stakes negotiations with OpenAI over the future of their strategic alliance, according to a report by the Financial Times published Wednesday. The talks have hit a stalemate over key disagreements, particularly regarding the size and structure of Microsoft’s future equity stake in the artificial intelligence company.

Sources familiar with the matter told the FT that Microsoft may pause or terminate discussions if no breakthrough is reached. In the meantime, Microsoft plans to lean on its existing commercial agreement, which guarantees access to OpenAI’s technologies, including its ChatGPT models, through 2030.

The situation comes amid increased tension between the two AI powerhouses. A separate Wall Street Journal report earlier this week revealed that OpenAI executives have considered accusing Microsoft of anticompetitive practices related to their ongoing partnership. Both companies are reportedly negotiating changes to Microsoft’s investment terms, including its future stake in OpenAI.

Despite the friction, both sides released a joint statement earlier this week affirming their intention to collaborate:

“Talks are ongoing, and we are optimistic we will continue to build together for years to come.”

Microsoft’s multi-billion dollar investment into OpenAI has positioned it as a central player in the AI boom, helping the company compete aggressively with rivals like Google and Amazon. The partnership has powered Microsoft’s integration of OpenAI models into products like Copilot in Microsoft 365 and Azure OpenAI Service.

Meanwhile, OpenAI is seeking approval from Microsoft—its dominant backer—to convert into a public-benefit corporation, a structural change the startup believes would facilitate greater capital raising flexibility.

The evolving rift highlights the complex interdependence between Big Tech firms and rapidly-scaling AI startups, raising questions about governance, control, and long-term alignment in the sector.

Meta’s $14.8 Billion Scale AI Deal Raises Regulatory Questions Amid AI Partnership Scrutiny

Meta’s $14.8 billion investment in data-labeling startup Scale AI, along with hiring its CEO, poses a test of the Trump administration’s stance on so-called “acquihire” deals—arrangements that some critics argue are used to bypass antitrust scrutiny.

The deal, announced Thursday, gives Meta a 49% nonvoting stake in Scale AI, which employs gig workers to manually label data and serves major clients including Meta’s rivals Microsoft and OpenAI. Because Meta does not gain a controlling stake, the transaction avoids mandatory U.S. antitrust review. Still, regulators could investigate if they suspect the deal was structured to sidestep rules or harm competition.

The structure aims to prevent Meta from cutting off competitors’ access to Scale’s services or gaining undue insight into rival operations. Despite this, Reuters reported that Alphabet’s Google has decided to sever ties with Scale following Meta’s investment, while other customers are reconsidering their relationships.

Scale AI stated its business remains strong and that it is committed to protecting customer data. Scale’s 28-year-old CEO Alexandr Wang will join Meta as part of the deal but will remain on Scale’s board with restricted access to sensitive information.

Experts say that while the Trump administration’s antitrust enforcers are cautious of big tech platforms, they generally want to avoid overregulating AI development. William Kovacic, competition law expert at George Washington University, noted regulators will watch these partnerships closely but might not intervene if they do not stifle competition.

Previous FTC inquiries into “acquihire” deals under the Biden administration—including Amazon’s hiring from AI startup Adept and Microsoft’s $650 million deal with Inflection AI—have so far resulted in no enforcement action.

Boston College Law professor David Olson highlighted Meta’s choice of a minority, nonvoting stake as a legal shield, though he acknowledged the FTC could still seek to review the deal.

The investment has drawn criticism from U.S. Senator Elizabeth Warren, who called for scrutiny to ensure Meta does not unlawfully suppress competition or increase monopoly power. Meta is already facing an FTC monopoly lawsuit, but whether regulators will challenge this specific investment remains unclear.

Separately, the U.S. Department of Justice’s antitrust division is probing whether Google’s partnership with chatbot maker Character.AI was structured to evade regulatory review and is seeking advance notice of Google’s AI investments as part of broader efforts to rein in the company’s dominance.

Reddit Misses Daily Visitor Estimates Due to Google Algorithm Change, Shares Drop

Reddit (RDDT.N) fell short of market expectations for daily active unique visitors in the fourth quarter, impacted by a change in Google’s search algorithm that reduced its visibility in search results. This announcement sent Reddit’s shares tumbling 15% in after-hours trading on Wednesday.

The San Francisco-based company, which launched its IPO in March 2024, saw its stock surge nearly five-fold last year. However, volatility with Google search at the end of the fourth quarter affected traffic from “logged-out users”—those who browse without signing in, CEO Steve Huffman said in a letter to shareholders. Despite the setback, Huffman noted that traffic from Google search has since recovered in the first quarter of 2025.

Reddit’s daily active unique visitors grew by 39% to 101.7 million for the quarter ending December 31, but this fell short of analysts’ average estimate of 103.3 million, according to LSEG data. Growth has also slowed sequentially.

“Reddit shares are down due to missed expectations on daily active users, but it’s not a reason to lose faith in the company,” said Jeremy Goldman, senior director of briefings at eMarketer. He added that Reddit’s international expansion and AI advancements could help it become a digital advertising powerhouse.

Reddit has been leveraging AI deals with Alphabet’s Google and Microsoft-backed OpenAI, as well as conversation placement ads—where brands can advertise directly within subreddit discussions. These factors have helped Reddit forecast first-quarter revenue between $360 million and $370 million, surpassing analysts’ average estimate of $358.1 million.

The company reported fourth-quarter revenue of $427.7 million, beating the $405.3 million estimate, largely driven by holiday season ad spending. Profit per share came in at 36 cents, surpassing the 25-cent estimate. Reddit’s global average revenue per user increased by 23% to $4.21.

Additionally, Huffman mentioned ongoing discussions for data licensing deals with major industry players.

Google has yet to respond to inquiries about its algorithm changes.