Barely weeks after introducing advertising into its flagship chatbot, OpenAI has already crossed $100 million in annual recurring revenue from the initiative, a figure that is drawing attention across Silicon Valley and the broader digital advertising industry.

The pilot, rolled out in January, is limited to free-tier users and ChatGPT Go subscribers in the United States. Even within that pool, exposure remains deliberately restricted. While roughly 85% of eligible users can be served ads, fewer than one in five encounter them on a daily basis — a constraint that underscores the company’s caution as it navigates a delicate trade-off between monetization and user trust.

Advertising inside ChatGPT follows a format that differs from traditional search or social media placements. Ads are positioned beneath responses, clearly labeled and visually separated from the model’s output. OpenAI has been explicit that commercial content does not shape answers generated by the system, a claim that goes to the heart of concerns about integrity in generative AI.

The company has also imposed category restrictions, barring ads from appearing alongside politically sensitive or health-related queries, and excluding users under 18 entirely. These guardrails reflect an awareness that conversational AI operates in a more intimate context than search engines or social feeds, where users often disclose personal or high-stakes information.

Early advertiser uptake has been strong. More than 600 brands are already participating, according to the company, signaling demand for access to what is effectively a new form of user engagement — one that captures intent in real time, often expressed in full sentences rather than keywords. For marketers, that shift offers the potential for more precise targeting, but it also raises questions about how far such targeting should go.

The speed at which revenue has accumulated is notable, particularly given the limited scale of the rollout. It suggests that pricing, or advertiser willingness to pay, is already robust — likely driven by the scarcity of inventory and the novelty of the format. In conventional digital advertising markets, scarcity tends to command a premium, especially when attached to high-engagement environments.

Even so, the rollout has not been without tension. Some advertisers have expressed frustration at the controlled pace, arguing that the limited availability of impressions constrains campaign reach. OpenAI’s response has been to emphasize experimentation over expansion.

“We’re in the early testing phase of ads in ChatGPT, and the goal right now is to learn and refine the experience for consumers before expanding it more broadly,” the company said. “We’re encouraged by early signals from users and participating brands, and continue to see strong interest from advertisers.”

The company is now extending tests beyond the United States, with early exploration underway in Canada, Australia, and New Zealand. The choice of markets, developed, English-speaking, and with mature advertising ecosystems, points to a measured internationalization strategy rather than an immediate global push.

OpenAI’s move comes as the economics of artificial intelligence grow more demanding. Training and operating large-scale models requires substantial computing infrastructure, and the cost of serving millions of queries daily continues to rise. Subscription products and enterprise licensing have so far underpinned revenue, but advertising introduces a potentially high-margin complement — provided it can scale without eroding user confidence.

The competitive backdrop is also shifting. Digital advertising remains dominated by Google and Meta, both of which have begun integrating generative AI into their own platforms. OpenAI’s entry into the market introduces a different paradigm: advertising embedded within dialogue rather than search results or social feeds.

Not all rivals agree with the approach. Anthropic has publicly criticized the move, using a high-profile advertising campaign to question the implications of blending AI assistance with commercial messaging. The critique comes off as part of a broader debate within the industry about whether conversational systems should remain insulated from the incentives that underpin traditional ad-driven platforms.

However, it is believed that OpenAI’s real challenge lies in sustaining a balance that has historically proven difficult in technology: extracting value from user attention without compromising the perceived neutrality of the product. The company’s insistence that ads are segregated from responses is designed to address that concern, but the longer-term test will be behavioral rather than technical.

The major test will likely be about users’ willingness to continue to trust the system as it becomes more commercialized. For now, the early figures suggest that advertisers are willing to bet on the format.

JPMorgan Chase is tightening its grip on how work gets done inside one of Wall Street’s largest technology operations, embedding artificial intelligence directly into how tens of thousands of engineers will be assessed, promoted—or left behind.

Internal documents reviewed by Business Insider show the bank has formally updated performance expectations for software and security engineers, making AI adoption a measurable requirement rather than a discretionary tool. The changes apply across its 65,000-strong Global Technology division, a workforce that underpins everything from trading systems to consumer banking platforms.

The overhaul is based on a directive that leaves little room for ambiguity.

“Demonstrate measurable improvement in code quality, speed and productivity through regular use of approved AI coding assist tools, contributing to the team’s overall efficiency targets,” one of the newly introduced goals states.

Engineers are also being asked to go further, beyond personal productivity, to reshape how work flows across the organization. Another directive instructs them to “engage in identifying, implementing and optimizing AI-driven automation opportunities within technology lifecycle management (TLM) processes to drive efficiency and support capacity unlock initiatives, ensuring all enhancements leverage current technology assets before considering new solutions.”

The language is not advisory. According to the internal materials, these objectives “will be added automatically and will appear by the end of March,” effectively standardizing AI usage as part of every engineer’s annual goals. Employees are expected to work with their managers to align individual targets with the new framework, ensuring that adoption is both tracked and enforced.

JPMorgan is already among the heaviest spenders on technology in global finance, with projected investments approaching $20 billion in 2026—well ahead of most competitors. The scale of that spending suggests the bank sees AI not simply as a productivity tool, but as a lever for structural cost reduction and operational speed at scale.

Inside the firm, the shift is already reshaping day-to-day dynamics.

Engineers say discussions about AI have intensified across teams, appearing in managerial briefings, internal communications, and performance dashboards. One such dashboard tracking GitHub Copilot usage reportedly drills down to individual employees, classifying them as “light,” “heavy,” or “non” users. It is an approach that turns tool adoption into a visible metric of engagement.

“There’s a lot of anxiety in the environment right now,” one longtime developer was quoted as saying, describing a workplace where AI usage is increasingly tied to perceptions of performance.

Another engineer said a manager made the expectation explicit during a recent meeting, telling staff that access to new AI tools comes with an “expectation” that output and delivery speed should show “a noticeable increase” quarter over quarter.

The bank’s expanding AI toolkit is reinforcing that expectation. A pilot rollout of Claude Code, developed by Anthropic, is expected as early as April, adding to a suite that already includes multiple models from Anthropic and OpenAI. The growing stack underpins a strategy of embedding AI across different layers of engineering work, from code generation to testing and documentation.

For many developers, the tools themselves are not in question. Several said AI has already proven useful in speeding up routine tasks and improving output. The unease stems from how tightly usage is being monitored—and what happens to those who fall short.

JPMorgan’s approach builds on a longer-standing culture of internal measurement. The bank has previously faced scrutiny over its Workforce Activity Data Utility, a system that tracked how employees spent their time, from the length of meetings to email drafting patterns. The new AI-focused metrics extend that philosophy into evaluating how work is produced, not just how it is scheduled.

At the same time, the firm is restructuring its broader performance management system. Employees will now be evaluated across two primary dimensions: “what you achieve,” focused on business outcomes, and “how you achieve it,” which includes adherence to internal behaviors and standards.

Under the revised framework, staff will be sorted into three categories: “stand out” for top performers, “achiever” for the majority, and “needs improvement” for those struggling to meet expectations. The system is designed to sharpen differentiation in a workforce where performance ratings have historically been more compressed.

AI adoption is being woven directly into those assessments. Internal materials list “data fluency” as a core competency, describing it as the ability to “develop and drive adoption of new tools or methodologies to leverage data in the flow of work.” Crucially, “rate of adoption” is cited as a measurable indicator of that skill, linking career progression to how quickly employees incorporate AI into their routines.

The bank has also made clear that performance tracking will be continuous. “You and your manager will use your objectives to track your progress during the year, recognize impact, and streamline your annual review,” one internal page states, reinforcing the role of ongoing measurement rather than end-of-year evaluation.

The implications extend beyond JPMorgan. Across corporate America, companies are beginning to treat AI proficiency as a baseline expectation, not a specialized skill. What is emerging is a new productivity benchmark—one where output is calibrated against what is possible with machine assistance, not just human effort.

At JPMorgan, that shift is being operationalized with precision. The bank is not just introducing new tools; it is redefining performance around them.

Senegal’s battle to reclaim its African Cup of Nations crown has shifted to the courtroom, with the country’s football authorities filing a formal appeal before the Court of Arbitration for Sport (CAS) in a case that is already stirring unease across the sport.

The dispute stems from the chaotic final in Rabat on January 18, where Senegal beat Morocco 1-0 after a brief abandonment. Players had walked off in protest against a penalty decision they deemed decisive. When they returned 14 minutes later and saw the match, few expected the result itself would later be erased.

But in a move that has since triggered widespread criticism, the appeals board of the Confederation of African Football (CAF) ruled that Senegal had effectively forfeited the match by leaving the field, awarding Morocco a 3-0 victory and, with it, the title.

CAS confirmed that Senegal’s appeal seeks not only to annul that ruling but to have the country formally reinstated as champions of Africa. The federation is also pressing for procedural safeguards, including a suspension of filing deadlines until CAF provides a fully reasoned decision — something Senegal’s legal team insists has yet to materialize.

At the core of the case lies a fundamental question about the boundaries of authority in football. Senegal’s lawyers argue that once a referee allows a match to resume and reach its conclusion, the result becomes sacrosanct. Any subsequent administrative reversal, they warn, risks weakening one of the sport’s most enduring principles — that outcomes are settled on the field of play.

Juan de Dios Crespo Perez, a senior member of Senegal’s legal team, did not temper his language. He described CAF’s ruling as “crude” and “irrational,” arguing that it runs counter to the laws of the game.

“This decision cannot even be considered a true sporting justice ruling – it is so crude, so absurd, so irrational,” he said. “It openly violates the laws of the ?game and the principle that refereeing decisions are final.”

His position has found resonance beyond Senegal’s borders, where former players, administrators, and analysts have questioned whether CAF’s disciplinary reach has crossed into territory traditionally reserved for match officials.

That criticism has been sustained and unusually broad. Across African football circles, the decision has been described as disproportionate, with commentators noting that walk-offs, while punishable, rarely result in the retroactive nullification of completed matches. Others have pointed to inconsistencies in past CAF disciplinary cases, arguing that sanctions have historically stopped short of rewriting results once play has resumed under a referee’s authority.

Legal observers have also raised concerns about due process. The absence of a detailed explanation accompanying the appeals board’s decision has drawn scrutiny, with some warning that opaque rulings risk eroding confidence in CAF’s governance structures. In similar cases globally, disciplinary bodies are expected to clearly articulate both the factual basis and regulatory framework underpinning their decisions, a standard Senegal insists has not been met.

Abdoulaye Fall, president of the Senegalese Football Federation, has framed the dispute in stark terms, calling the ruling an “administrative robbery” and pledging a sustained legal campaign. His rhetoric reflects the domestic mood, where the government has already called for an inquiry into how the title was stripped.

The legal team assembled by Senegal underscores the seriousness of the challenge. Drawing from jurisdictions including Switzerland, Spain, France, and Senegal, the group is preparing for what could become a defining case in sports arbitration. Seydou Diagne, another lawyer on the team, warned that the implications extend well beyond this single final.

“If CAS let this situation happen, the winner of the next World Cup could be decided within a lawyers’ firm,” he said.

CAS jurisprudence has historically been cautious about interfering with refereeing decisions, recognizing the autonomy of match officials as a cornerstone of the sport. Senegal’s argument leans heavily on that tradition, positioning CAF’s ruling as an outlier that risks redrawing those boundaries.

CAF president Patrice Motsepe has defended the organization’s stance, insisting that fairness and uniformity guided the decision-making process. Yet even within administrative circles, there is quiet acknowledgement that the case could test the limits of CAF’s disciplinary authority.

While standard CAS procedures can stretch over nine to 12 months, Senegal is pushing for an accelerated process, aware that the symbolic and commercial weight of the AFCON title diminishes with prolonged uncertainty. Any fast-tracking would require agreement from CAF and Morocco, adding another layer of complexity.

“Such a procedure usually lasts nine to 12 months, but we want it to go faster. However, all parties must agree to it,” Serge Vittoz, part of a ?six-lawyer legal team in Paris, said.

In the interim, Senegal’s position remains unchanged. Officials insist the team is, in their view, still African champions. There are even suggestions that the trophy could be presented to supporters during upcoming fixtures in Europe, a move that would underline the federation’s refusal to recognize CAF’s decision.

What began as a disputed penalty in Rabat has evolved into a test case for football governance. The case is expected to shape how far administrative bodies can go in overturning results, and whether the final whistle truly marks the end of a match.