OpenAI announced on February 3, 2026, that Dylan Scandinaro would become its Head of Preparedness after working on AI-safety efforts at Anthropic. The position had been advertised in December 2025 with base pay of up to $555,000 a year, plus equity. That is the job listing’s ceiling—not verified evidence of Scandinaro’s personal salary.
What is confirmed about Dylan Scandinaro’s move?
OpenAI named Scandinaro its Head of Preparedness in an announcement reported by Bloomberg-sourced coverage. Public reporting and professional posts identify him as a former member of Anthropic’s AI-safety team. Scandinaro described joining OpenAI’s preparedness effort and expressed appreciation for his time at Anthropic.
The available reporting does not establish his exact former title, tenure, compensation, or which Anthropic projects he worked on. “Former Anthropic AI-safety researcher” is therefore more precise than calling him Anthropic’s safety chief.
Does Scandinaro earn $555,000?
Not necessarily. OpenAI advertised the Head of Preparedness position with base compensation of up to $555,000 annually, plus equity. A maximum listed salary is not a guaranteed offer, and no public source cited here confirms Scandinaro’s individual package.
Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Clear out junk files and repair common Windows errors3Fix the driver behind crashes, sound loss and screen glitches#1 Best Overall
- OpenAI develops a unified AI platform designed to act as a proactive assistant for daily life. OpenAI provides advanced reasoning, multimodal search, and agentic workflows that simplify complex tasks to expand what individuals and teams can build.
- OpenAI is for entrepreneurs, teachers and students, artists and scientific researchers. OpenAI is for those building AI-native projects, creating art, or working with data workflows who need a digital agent to accelerate learning and project completion.
- Lightweight, Classic fit, Double-needle sleeve and bottom hem
| Compensation element | What is established |
|---|---|
| Base salary | Up to $555,000 a year in the advertised listing |
| Equity | Included in the listing; amount and vesting are not stated |
| Total compensation | Not verified |
| Scandinaro’s actual salary | Not verified |
| Bonus, benefits, location and level | Not verified |
Fortune’s account of the listing describes the same upper-end base-pay figure and the role’s unusually broad risk remit: Fortune. Bloomberg News also circulated the “listed at $555,000” wording in a LinkedIn post.
What does OpenAI’s Head of Preparedness do?
Preparedness is not ordinary content moderation. The job concerns whether frontier-model capabilities could create severe real-world harm and whether OpenAI has controls ready before deployment. Recruitment coverage says the role involves assessing emerging capabilities, building mitigations and preparing response plans.
Rank #2
- OpenAI develops a unified AI platform designed to act as a proactive assistant for daily life. OpenAI provides advanced reasoning, multimodal search, and agentic workflows that simplify complex tasks to expand what individuals and teams can build.
- OpenAI is for entrepreneurs, teachers and students, artists and scientific researchers. OpenAI is for those building AI-native projects, creating art, or working with data workflows who need a digital agent to accelerate learning and project completion.
- Lightweight, Classic fit, Double-needle sleeve and bottom hem
Core responsibilities
- Track capabilities that could enable high-consequence misuse.
- Evaluate risks before and around deployment of advanced models.
- Develop technical safeguards, monitoring and incident-response plans.
- Coordinate research, security and operational teams when a risk threshold is reached.
- Address both foreseeable misuse and uncertain scenarios involving more autonomous or self-improving systems.
Risk areas named in the recruitment material
- Cybersecurity threats
- Biological risks
- Mental-health effects
- Systems capable of improving their own performance
TechCrunch’s report on the listing describes a team created in 2023 to study risks ranging from immediate abuse, such as phishing, to longer-horizon catastrophic scenarios. The remit is therefore broader than screening user prompts or removing harmful content.
Why is the position a “hot seat”?
“Hot seat” is an analytical description, not OpenAI’s formal job title. The pressure comes from having to make uncertain technical judgments inside a company whose products and revenue depend on shipping quickly.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Rank #3
- OpenAI develops a unified AI platform designed to act as a proactive assistant for daily life. OpenAI provides advanced reasoning, multimodal search, and agentic workflows that simplify complex tasks to expand what individuals and teams can build.
- OpenAI is for entrepreneurs, teachers and students, artists and scientific researchers. OpenAI is for those building AI-native projects, creating art, or working with data workflows who need a digital agent to accelerate learning and project completion.
- 8.5 oz, Classic fit, Twill-taped neck
- Speed versus safety: Additional evaluations or safeguards can delay, narrow or condition a launch.
- Known versus unknown risks: Testing can target foreseeable misuse, but capability jumps and emergent behavior are difficult to predict.
- Expertise versus authority: A preparedness leader can identify a danger without necessarily having unilateral power to stop a release.
- Secrecy versus scrutiny: Frontier labs may limit disclosure of capabilities while asking outsiders to trust their internal controls.
Sam Altman characterized the work as stressful and said the successful candidate would enter the “deep end” quickly. He also said OpenAI would soon be working with “extremely powerful models,” according to WinBuzzer’s account. Those remarks signal urgency; they do not establish that OpenAI has achieved AGI, that a particular model launch is imminent or that self-improving systems are already operating in production.
Why does the Anthropic connection matter?
Anthropic has built much of its public identity around AI safety, constitutional AI and model safeguards. Recruiting someone from its safety organization gives OpenAI experience from a direct rival and shows that preparedness expertise has become a scarce labor-market asset alongside model researchers, infrastructure engineers and product leaders. Business Standard’s report places the move in that competitive context.
Rank #4
- OpenAI develops a unified AI platform designed to act as a proactive assistant for daily life. OpenAI provides advanced reasoning, multimodal search, and agentic workflows that simplify complex tasks to expand what individuals and teams can build.
- OpenAI is for entrepreneurs, teachers and students, artists and scientific researchers. OpenAI is for those building AI-native projects, creating art, or working with data workflows who need a digital agent to accelerate learning and project completion.
- Lightweight, Classic fit, Double-needle sleeve and bottom hem
The hire does not prove that Anthropic is inherently safer, that OpenAI is inherently unsafe, or that Scandinaro speaks for Anthropic. There is also no verified reporting here of an objection, noncompete dispute or strategic disagreement between Scandinaro and his former employer.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What is the history behind OpenAI’s preparedness post?
OpenAI created its preparedness team in 2023. Earlier reporting said Aleksander Madry, an earlier Head of Preparedness, was later reassigned to work focused on AI reasoning less than a year after the role was created. That turnover helps explain why commentators describe the post as demanding, but it is not evidence that the team is dysfunctional. The organizational history is summarized in TechCrunch’s coverage.
Free tools Windows power users keep installed
One-click scans. No signup required.
Best Value
- OpenAI develops a unified AI platform designed to act as a proactive assistant for daily life. OpenAI provides advanced reasoning, multimodal search, and agentic workflows that simplify complex tasks to expand what individuals and teams can build.
- OpenAI is for entrepreneurs, teachers and students, artists and scientific researchers. OpenAI is for those building AI-native projects, creating art, or working with data workflows who need a digital agent to accelerate learning and project completion.
- Lightweight, Classic fit, Double-needle sleeve and bottom hem
What the appointment does—and does not—show
What it suggests
The senior title, executive attention and unusually high advertised salary suggest OpenAI considers preparedness strategically important and is willing to pay for scarce expertise. They also indicate that safety governance is part of the competition for frontier-AI talent, not a peripheral compliance function.
What it does not establish
- That Scandinaro receives $555,000 in base pay.
- That his equity or total compensation is known.
- That he can independently delay or veto product releases.
- That a specific upcoming system prompted the hire.
- That OpenAI’s safeguards have been proven effective.
- That an imminent breakthrough or autonomous system is confirmed.
Scandinaro’s own recruitment communication is available on LinkedIn; a contemporaneous reproduction appears at Gigazine. Neither source turns the advertised compensation ceiling into a disclosed personal contract.
Bottom line
OpenAI hired a former Anthropic AI-safety researcher to lead preparedness work covering severe risks from increasingly capable models. The role was advertised at up to $555,000 in base salary plus equity, but Scandinaro’s actual pay is not public. The appointment is significant because it places scarce safety expertise at the center of a rivalrous frontier-model market; it is not, by itself, proof that OpenAI has solved safety governance or that a particular powerful system is about to launch.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
