AI Leadership Weekly

Top Stories

US Senate removes “AI moratorium” from bill
A controversial plan to block US states from regulating artificial intelligence for the next decade has been decisively scrapped from a major Senate budget bill, following widespread bipartisan opposition. The so-called “AI moratorium,” originally pushed by Texas Senator Ted Cruz and strongly endorsed by several high-profile Silicon Valley figures, would have stopped states from passing their own AI rules until at least the mid-2030s. After some last-minute negotiating and a dramatic reversal from co-sponsor Senator Marsha Blackburn, the Senate rejected the measure by 99 votes to one.

• Tech leaders back the moratorium. OpenAI’s Sam Altman, Anduril’s Palmer Luckey, and VC heavyweight Marc Andreessen supported the moratorium, arguing a patchwork of state-level AI laws would “stifle innovation” and create confusion for companies trying to deploy or develop AI in the US.

• Bipartisan push-back against the ban. Opposition quickly coalesced across party lines, with critics warning that “the ban on state regulation would harm consumers, and let powerful AI companies operate with little oversight.” Many legislators saw the moratorium as handing too much unchecked power to tech giants, at a time when scrutiny of big tech is already intensifying.

• Short-lived compromises and swift rejection. Senator Blackburn initially negotiated the ban down from 10 years to five, but then withdrew her support. By Monday, bipartisan sentiment had shifted so strongly that the Senate overwhelmingly voted to remove the provision.

Why it matters: The near-unanimous Senate rejection underscores a growing appetite on Capitol Hill for stronger oversight of the AI sector, even against the wishes of tech industry titans. With federal AI regulation still uncertain, states are likely to move ahead with their own laws with a clear message to tech executives: self-regulation is not enough.

97% of AI users pay nothing
Despite a surge in global AI adoption, most of the nearly two billion users are not paying a penny for those tools, according to a new report from Menlo Ventures. While AI companies have invested vast sums into infrastructure, just 3% of the estimated addressable market has translated into actual revenue. The sector now faces real questions about how (and when) all those free users might turn into profit.

• A massive opportunity remains untapped. With 1.8 billion AI users worldwide—and a staggering third using AI tools daily—the potential revenue is enormous. Menlo Ventures calculates that if every user paid $20 per month, the AI consumer market could be worth $432 billion, not the paltry $12 billion seen today.

• Turning free users into paying customers. Analyst Shawn Carolan suggests unlocking features like memory assistance and transaction capabilities to entice users into premium subscriptions, rather than watching them switch to the next free tool once their credits dry up.

• Looking beyond subscriptions. Carolan reckons the bigger “opportunity” might be monetising free users with targeted advertising and affiliate fees. With AI-driven personalisation, he claims, ads can be both more relevant for users and more lucrative for tech firms.

Why it matters: The AI boom has a glaring business problem: almost no one pays for these tools yet, despite staggering investment. If AI firms don’t crack the code on monetisation—either through subscriptions, advertising, or new models—they risk building the world’s most expensive “free lunch.”

Sam Altman downplays Meta’s talent poach
OpenAI CEO Sam Altman has fired back at Meta’s aggressive AI talent poaching, telling his team that “missionaries will beat mercenaries” and hinting that Meta’s approach may lead to “very deep cultural problems.” In a leaked memo, Altman took direct aim at Mark Zuckerberg’s recent hiring spree—which included several ex-OpenAI engineers—and sought to rally the troops around OpenAI’s mission and unique culture.

• Meta’s hiring frenzy prompts backlash. Zuckerberg just unveiled Meta’s new superintelligence team, mostly staffed with high-profile hires from competitors, including OpenAI. The exodus prompted OpenAI’s chief research officer to say it felt as if “someone has broken into our home and stolen something.”

• Culture vs. cash debate heats up. Altman claims Meta “didn’t get their top people and had to go quite far down their list,” and makes the case that OpenAI’s sense of mission and its pursuit of artificial general intelligence for the greater good are what really set it apart. In his words: “Missionaries will beat mercenaries.”

• OpenAI responds with compensation review and team pride. Acknowledging the lure of massive offers, OpenAI is ramping up a review of staff pay but insists that culture remains its core competitive edge. Staff chimed in to double down, calling OpenAI a “magical cradle of innovation” and praising its “quirky and weird” personality as a strength.

The bottom line: The AI talent war is about more than sky-high salaries. For top researchers, a compelling mission and a strong culture might still outweigh even the most lucrative offers, although OpenAI knows it can’t ignore the money either. As Altman puts it, OpenAI plans to be around for the long haul, “day after day, year after year,” as rivals come and go.

In Brief

• Huawei is open sourcing models from Pangu series

• New Microsoft diagnostic system 4 times more accurate than human doctors

• Men turning to ChatGPT for mental health support

• SciArena evaluates foundation models in scientific literature tasks

Market Trends

Top AI engineers lured with $10m+ salaries
Big Tech is locked in a frenzied war for AI talent, with top engineers and researchers now commanding superstar salaries that would make most CEOs blush. Word is that Meta has dangled $100 million sign-on bonuses in front of OpenAI’s best, prompting OpenAI to urgently review its own compensation and dream up “creative ways to recognise and reward top talent.” The scramble is reaching fever pitch as companies like Meta, Google, and OpenAI race to claim the AI high ground.

• Salaries soar to eye-watering heights. In Silicon Valley, salaries for senior AI engineers have ballooned, with many taking home $3–7 million a year, and a lucky few even cracking $10 million. For reference, regular software engineers are lucky to get even a third of that.

• Meta lights a fire under the competition. Mark Zuckerberg’s Meta, eager to close the gap after an underwhelming Llama 4 launch, has been particularly aggressive, dangling jaw-dropping offers and signifying just how cutthroat the competition has become. OpenAI is giving staff a week off to resist Meta’s overtures, while execs work "around the clock" to keep staff on side.

• Not just about the money. Recruiters say that reputation, research freedom, and a sense of purpose still matter, and might actually tip the scales against the biggest pay packets. As Aleph Alpha’s CEO puts it, “belief in a mission” is something money can’t buy.

The bottom line: The AI wage war shows no sign of slowing and it’s no longer just about who pays best, but who offers the best mission, team, and impact. For startups and non-giants, creative offers and a compelling company culture may be the last real hope for luring (and keeping) tomorrow’s AI superstars.

Siri’s future may run on OpenAI or Anthropic, not Apple tech
Apple is reportedly weighing a major shake-up to its flagship voice assistant Siri, considering deals with Anthropic or OpenAI to inject their cutting-edge AI models into future versions. This is a dramatic pivot for the iPhone maker, which has traditionally insisted on building its own technology in-house, but now appears to be conceding ground in the generative AI race. If the move goes ahead, Siri could soon run on either Anthropic’s Claude or OpenAI’s ChatGPT, rather than Apple’s “Apple Foundation Models”.

• Admitting a generative AI gap. Behind the scenes, executives are said to recognise Apple has fallen behind rivals in AI, with one source calling the shift “a monumental reversal” and a public sign of Apple “struggling to compete”. The company’s earlier plan for a new Siri, powered solely by its own models and slated for 2026, is now in doubt.

• Tensions and talent turmoil. The prospect of third-party AI powering Siri has deeply unsettled Apple’s internal AI team, with some engineers threatening to leave for far more lavish compensation elsewhere. As one open letter internally framed it, the move may “create the perception that they are to blame for the company’s AI shortcomings”. Staff departures and poaching by firms like Meta have further destabilised morale.

• Negotiations and privacy priorities. Apple is still hammering out potential terms with Anthropic and OpenAI, including whether custom versions of the models can run on Apple’s own cloud hardware, a compromise designed to protect user privacy. But the financial cost could be steep, with Anthropic reportedly seeking a rapidly escalating multi-billion-dollar annual fee.

The big picture: Apple’s possible U-turn on AI reveals just how fast the industry is moving, and how even the world’s most secretive and self-reliant tech giant feels pressure to keep up. For entrepreneurs and executives alike, it’s a sharp reminder that building AI in-house isn’t always enough. Sometimes it’s adapt or get left behind.

Cloudflare proposes “pay to scrape” scheme
A pivotal shift is underway in the relationship between content creators and AI giants, with Cloudflare and a coalition of global publishers now blocking AI crawlers unless they pay for the privilege. Announced as “Content Independence Day,” this move signals the end of an era where creators exchanged their content for web traffic, and a response to fears that AI is “stripmining” the web while sending creators little to no value in return.

• The old deal, now broken. For decades, the pact was simple: Google (and later others) indexed your work and sent you visitor traffic. But an increasing share of user questions are now answered directly by AI systems, so “it’s 750 times more difficult to get traffic from OpenAI than the Google of old,” and basically impossible for some.

• AI threatens the economics of online content. Instead of rewarding originality and quality, AI tools often serve up derivative answers, eating away at the ad, subscription, and visibility models that once sustained the free web.

• Publishers push back with new terms. Cloudflare and its partners are drawing a line: no more AI crawling without direct compensation for creators. They’re also working towards a new marketplace that values content based on its unique contribution to knowledge, not blunt traffic numbers.

Why it matters: How we reward creators will define the future of the web. As AI becomes the new gateway to information, whether or not content makers get a fair shake could determine if the next web generation is flush with valuable, original work, or simply more algorithmic mush.

Tools and Resources

UniDeck

Generate dashboards personalised to your needs.

Cora

Cora is like your "AI chief of staff".

AutoLocalise

Use AI to automatically localise your app.

Hit reply to let us know which of these stories you found the most important or surprising! And, if you’ve stumbled across an interesting link/tweet/news story of your own, send it our way at [email protected] It might just end up in the next issue!

Thanks for reading. Stay tuned for the next AI Leadership Weekly!

Brought to you by Data Wave your AI and Data Team as a Subscription.
Work with seasoned technology leaders who have taken Startups to IPO and led large transformation programmes.