AI + Careers Briefing — November 1, 2026

Today's two stories are about the terms on which people use AI at work. One is about cost: ChatGPT is moving its paid plans to a shared, metered token budget. The other is about conduct: a Brazilian court penalized two lawyers for hiding an instruction aimed at an AI reader. For job seekers, the takeaway is the same from both directions. Use AI deliberately, know what it costs you, and never try to game the machines on the other side of an application.

Is ChatGPT's flat-rate pricing ending?

Partly. Starting in December, ChatGPT's Plus, Pro and Business plans will draw on one shared token budget covering both ChatGPT and Codex, as OpenAI merges the app's Chat and Work modes into a single interface, The Decoder reported on November 1.

Tokens measure the information a model reads and writes, so complex tasks use up the budget faster. According to the report, users will see their remaining balance and reset times in Settings; at the limit, they can wait for a reset, upgrade or buy credits. Conversations using Instant models will not count against the budget. The Decoder's own assessment is that heavy use on lower-priced plans will likely mean lower-quality output — that is the outlet's view, not an OpenAI statement. A third-party breakdown from laozhang.ai, dated October 8, lists Plus at $20 a month and says credits apply only to Work, Codex and Office add-ins.

Why it matters for job seekers: Treat AI as a small budget, not an endless utility. Long, multi-step sessions — tailoring dozens of applications at once, or iterating on a portfolio project with an agent-style tool — are the most likely to drain a shared pool. Batch heavy work like resume rewrites and mock interviews, and keep quick lookups in lighter modes. If you're employed, check whether your employer pays for a Business plan before buying your own. Our guides on building an AI productivity stack and job searching in the AI era cover which tasks justify paid tools. Plan details can change before December, so confirm terms with OpenAI directly.

Can hiding an instruction for AI get you penalized?

Yes. A Brazilian labor court sanctioned two lawyers for concealing an instruction aimed at artificial intelligence inside a court petition, with a penalty of 10 percent of the claim's value, according to eDiscovery Today, which first reported the ruling in June. Reporting describes it as the first case of its kind.

The text was set in white font on a white background, invisible to a human reader. Jetstream Security reports it told an AI system to respond to the petition superficially and not challenge the documents. The filing went through Galileu, an AI tool used by Brazil's National Justice Council, which detected the concealed instruction and blocked processing, per Jetstream. Minnesota Lawyer reports the judge held both plaintiff-side attorneys jointly liable and referred the matter to Brazil's bar association and a regional inspector general's office. eDiscovery Today notes the decision is subject to appeal and that the lawyers denied wrongdoing, saying the instruction was meant to protect their client.

Why it matters for job seekers: This was a court filing, not a hiring process, and none of the sources connects it to resumes. But hiring teams increasingly run applications through AI screening, and slipping invisible text into a document to steer a machine reader is the same trick. We found no source measuring how often applicants try it, so we won't claim it's common. Our read: a hidden instruction in a resume is a trust risk, and institutions are starting to treat concealed AI manipulation as misconduct — and, as here, the AI itself can flag it. Match your resume's real wording to the job description instead; our guide to optimizing your resume for ATS and AI screeners covers what legitimately helps.

What to watch

• Whether OpenAI's December rollout matches the reported token-budget design, and what the published limits per plan turn out to be.
• Whether the Brazilian ruling survives appeal, and whether other courts or employers set explicit rules on hidden AI instructions in documents.