AI + Careers Briefing — October 20, 2026

Today's stories share one idea: AI's impact depends less on the technology than on who controls the tasks, the tools and the rules around it. History suggests routine clerical work, much of it held by women, is most exposed. Microsoft now says the AI agents companies deploy need watching like any untrusted insider. Goldman Sachs says the real limit on AI's growth may be electricity, not software. And New York City's mayor argues local officials should fill the gap left by thin federal rules. For job seekers, the useful question is which parts of your work involve judgment and oversight, because those are the parts each story suggests will stay in human hands.

Are women at greater risk of AI job losses?

Women are likely to be among the most affected by AI because clerical and administrative work is the kind of routine work AI targets, according to a CNBC analysis of how past technologies reshaped women's jobs.

Ariane Hegewisch, a senior research fellow at the Institute for Women's Policy Research, told CNBC that AI is focused on routine tasks, numbers and finding patterns, and that clerical and administrative jobs are ones many women hold. "It's still 90% of secretaries and clerical workers who are women." CNBC also cited the International Labour Organization, which concluded in March that nearly 1 in 10 women in high-income countries work in occupations at the highest risk from AI.

CNBC points to two precedents. In the 1940s, nearly 95% of U.S. telephone operators were female, and an estimated 75,000 operator jobs vanished in the decades after automated switches spread. Economists James Feigenbaum and Daniel P. Gross found many operators ended up in lower-paying jobs or left the workforce, but the next generation of women found work in other fields, often of similar quality. The personal computer hit a similar group: the 1980 census counted nearly 4.7 million secretaries, stenographers and typists, more than 98% of them women. CNBC notes a 2022 paper by Marcus Dillender and Eliza Forsythe found software adoption in office and administrative work brought a 1% employment decline in the sector and a 3% wage increase for college graduates.

Why it matters for job seekers: The history in this report suggests exposure depends on the tasks in a role, not the job title. If your work is mostly routine processing, the pattern CNBC describes is that tasks shift to others or to software, while people who learned the new tools often stayed valuable. Reviewing our research on this, we would map which of your daily tasks are routine and which need judgment, then build proof of the second kind. Our guide to AI-proof career skills is a starting point. CNBC adds a caution: Hegewisch said the outlook is unclear because "AI can learn."

Source: CNBC

Why is Microsoft treating AI agents like insider threats?

Microsoft introduced new security features on Monday, October 19, built on the premise that the AI agents a company deploys can behave like a rogue employee, according to TechCrunch.

The headline addition is Guardian Agents, a feature intended to monitor the AI agents a company uses. TechCrunch reports that Microsoft's security business chief Hayete Gallot wrote, "Unfortunately, frontier models, regardless of their origin, cannot be trusted," adding that the reason is unpredictability rather than malice. Microsoft CEO Satya Nadella has made a similar argument, writing, "We need to separate the supply of intelligence from the authority over it."

Microsoft also added an agent detection feature to Defender that can find up to 60 types of agents employees bring into the workplace, including Claude Code, Codex and OpenClaw. New controls limit agents' access to data and systems and adjust permissions by task, including requiring human approval if an agent appears to be making risky decisions. TechCrunch notes the approach is also self-serving, since Microsoft wants enterprises to buy agent security from it.

Why it matters for job seekers: If employers treat agents as untrusted insiders, someone has to watch them, and the product features above include human approval steps. That is our read, not something the article states. Roles in security, IT and compliance may increasingly involve knowing how agents are permissioned and monitored. If you use agents at work, expect scrutiny of what they can access. Our advice is to learn those controls and describe that skill plainly on your resume. For what AI skills are worth in pay, see our look at the AI skills salary premium.

Source: TechCrunch

Is power the biggest bottleneck for the AI race?

Goldman Sachs says it is: in a new report shared exclusively with Axios, the bank warns that energy bottlenecks are set to stall the AI race and that power is the biggest risk to trillions of dollars in spending.

Goldman expects global data center power demand to surge 170% between 2025 and 2030, with more than 60% of that growth coming from the U.S., according to the report. The bank estimates roughly $7.6 trillion in capital spending between 2026 and 2031 across computing, data centers and power infrastructure. "Power is probably the forever bottleneck," Jason Tofsky, Goldman's global head of digital infrastructure investment banking, noted in the report.

Building supply is slow. Axios reports the U.S. has roughly 2,600 gigawatts of proposed power generation and storage projects awaiting grid connections, more than twice the grid's existing installed capacity, though many projects may never materialize. The median time from a grid connection application to commercial operation is roughly five years. Since January, 86 projects have been canceled or significantly delayed, including 76 directly linked to local opposition, according to Goldman. Utility capital plans are expected to reach $1.3 trillion between 2026 and 2030.

Why it matters for job seekers: The report does not discuss hiring, so this part is our analysis. If power and grid capacity limit how fast AI can grow, the spending the report describes lands on utilities and power infrastructure as well as on software. Job seekers weighing where AI-adjacent demand may sit could look at energy, grid and construction-related roles alongside model-building ones. The report also describes a longer, less predictable buildout, so timing is uncertain. For how employers weigh transferable skills, see our guide to skills-based hiring.

Source: Axios

What did NYC Mayor Mamdani say about AI and federal regulation?

New York City Mayor Zohran Mamdani said at a Manhattan event on Monday that the federal government has left a gap on AI oversight, and he urged New Yorkers to resist the sense that AI is inevitable, according to TechCrunch.

Mamdani spoke at The People's AI Assembly, an event hosted by the AI Now Institute, at Webster Hall, in conversation with activist and "Broad City" star Ilana Glazer. TechCrunch reported that he encouraged New Yorkers to resist the feeling that AI is coming for their jobs, their kids' minds, and their dignity. He argued that the technology's rapid rollout is not inevitable and that public officials should hold AI companies accountable. "Nothing has to be," he said.

He criticized what he called the "complete abdication of federal law" on AI. TechCrunch noted that President Donald Trump's Superintelligence Task Force, announced earlier this month, largely focuses on letting AI companies self-regulate rather than creating new federal restrictions.

TechCrunch pointed to two city actions. Last month, Mamdani ordered a ban on AI use in elementary and middle schools. Over the summer, he also introduced a policy that would require landlords, brokers, and real estate platforms to clearly disclose when listing photos or videos are made with AI. The article does not mention any city rules on AI in hiring.

Why it matters for job seekers: The article describes city officials stepping in where federal rules are thin, but it does not cite any NYC policy on AI hiring tools or workplace AI, so nothing changes yet for applicants. The takeaway is that AI rules may increasingly vary by city. Check local requirements where you apply, and see our guide on how to job search in the AI era for practical steps.

Source: TechCrunch

What to watch

• Whether more employers add human-approval steps for AI agents, and how that shows up in security, IT and compliance job descriptions.
• Utility and grid capital plans as a signal of where AI-related spending, and possibly hiring, lands next.
• Whether New York City extends its AI disclosure approach beyond schools and real estate listings.