Blog

<figure id="figure-t1ms50l1dkm6mjdlph"data-full-width><img id="t1ms50l1dkm6mjdlph" src="https://cdn.durable.co/blocks/18Qn5AYdtmZX5MK8Cyem8j6u5KtkxT4z3Y5nqCSJtFaHY0nvEeYFfQTrdQRciovQ.jpg" alt="" title="" class="object-cover bg-gray-100 " style="aspect-ratio: 16/9; object-position: center center;" data-corners="default" data-aspect="16:9" data-caption="" data-positionX="undefined" data-positionY="undefined" data-fullWidth="true"/></figure>Oh lookie here… this week another AI model is making waves in the tech world! Alibaba released a new version of its Qwen 2.5 model claiming it surpasses DeepSeek. But before we raise a glass or start handing out high fives, we need to look harder at DeepSeek. DeepSeek's big splash last week turned out to be less "breakthrough innovation" and more "breaking through every safety barrier in sight." As the new kid on the block, DeepSeek's R1 apparently never met a content filter it couldn't ignore.Recent <a data-mce-href="https://www.wired.com/story/deepseeks-ai-jailbreak-prompt-injection-attacks/?utm_source=nl&amp;utm_brand=wired&amp;utm_mailing=WIR_Daily_020125&amp;utm_campaign=aud-dev&amp;utm_medium=email&amp;utm_content=WIR_Daily_020125&amp;bxid=5c95706e3f92a40ad9b85203&amp;cndid=14654347&amp;hasha=9108ebf06b63664ffc46d71249c3f578&amp;hashc=ce5e925cc9d5f4eb915cbde78db518fc659a22ce952ac4901de7edd4308e036d&amp;esrc=MARTECH_ORDERFORM&amp;utm_term=WIR_Daily_Active" href="https://www.wired.com/story/deepseeks-ai-jailbreak-prompt-injection-attacks/?utm_source=nl&amp;utm_brand=wired&amp;utm_mailing=WIR_Daily_020125&amp;utm_campaign=aud-dev&amp;utm_medium=email&amp;utm_content=WIR_Daily_020125&amp;bxid=5c95706e3f92a40ad9b85203&amp;cndid=14654347&amp;hasha=9108ebf06b63664ffc46d71249c3f578&amp;hashc=ce5e925cc9d5f4eb915cbde78db518fc659a22ce952ac4901de7edd4308e036d&amp;esrc=MARTECH_ORDERFORM&amp;utm_term=WIR_Daily_Active" target="_blank" rel="noopener">research</a> from Cisco and the University of Pennsylvania revealed concerning findings. The situation is comparable to those horror movie teenagers who, while being chased by a masked killer, run upstairs instead of out the front door, or decide to "split up to cover more ground," or better yet, investigate the strange noise in the dark basement while saying "Hello? Is anyone there?" That's basically what happened when researchers tested DeepSeek's model with 50 different malicious prompts. It failed to stop a single one. It's likened to watching someone in a horror film walk into a cabin in the woods that has a sign saying "DANGER: DEFINITELY HAUNTED, ALL WHO ENTER WILL DIE" written in what appears to be fresh blood on the door, and thinking, "Yep, this seems like a cozy place to spend the night!"<h3>When "Cheaper" Comes at a Cost</h3>DeepSeek burst onto the scene promising a more affordable alternative to the big players like OpenAI. But as DJ Sampath from Cisco points out, there might be a reason for those cost savings – and it's not because they found a magical efficiency breakthrough. It seems they may have skipped the whole "robust security measures" chapter listed in any development handbook.Think of it like buying a car without airbags, seatbelts or brakes. Sure, it's cheaper, but do you really want to take that ride?<h3>The Jailbreak Express</h3>Here's where things get even juicier. Remember those clever "D.A.N." (do anything now) prompts that used to work on ChatGPT? That's child's play compared to what's possible with DeepSeek's R1. Adversa AI's analysis shows that the model's safety restrictions can be bypassed faster than a Labrador Retriever finding any body of water, including mud puddles to play in.And we're not talking about sophisticated, never-before-seen attacks here. These are the equivalent of using "password123" as your security key – basic, well-known vulnerabilities that have been making the rounds for years.<h3>Why Should We Care?</h3>Because AI isn't just playing around in sandboxes anymore. When these models get integrated into important complex systems, those "harmless" jailbreaks can cascade into serious business risks faster than you can say "breach nightmare."<h3>The Wake-Up Call</h3>Here's the thing: we can't afford to treat AI safety like an optional feature. It's not only reckless but dangerous to develop without security being at the cornerstone and top of mind. This was never truer during my time at AWS when Werner Vogels always said: "Everything will fail, and everyone's job is security." When companies rush to market with inadequate safeguards, they're not just cutting corners – they're potentially unleashing tools that could be weaponized for everything from spreading misinformation to creating harmful content.<h3>The Regulatory Void</h3>Here's where things get really interesting (and by interesting, I mean concerning). While established players like OpenAI operate under increasing scrutiny and evolving regulatory frameworks, newcomers like DeepSeek seem to be operating in a regulatory wild west. It's like watching someone open a restaurant without health inspections – sure, the food might be cheaper, but do you really want to risk it? Those who've ever had the very personal experience of food poisoning know exactly what that means.The continued absence of comprehensive international AI safety regulations means we're essentially trusting AI companies to police themselves. And as this case demonstrates, not everyone takes that responsibility as seriously as they should.<h3>The Ethics Gap</h3>Let's be real: ethical AI development isn't just about preventing harmful content – it's about building responsible technology that serves EVERYBODY while protecting users and society. When companies skip safety checks, they're not just cutting corners; they're potentially undermining the entire foundation of AI development.<h3>What Can We Do About It?</h3>1. Do Your Homework: If you're considering implementing AI solutions, make safety features and ethical considerations non-negotiable parts of your evaluation criteria.2. Require Transparency: We cannot continue to expect someone else to do it or do the right thing. While we are developing new tech at the speed of light, we need to rapidly ramp up and actually implement AI safety measures, independent security audits, and ethical guidelines. Let's stop talking about it and do it.3. Stay Informed: Keep up with security research, regulatory developments, and understand what makes an AI model truly "production-ready," before you deploy in your enterprise.I love the way Alex Polyakov from Adversa AI put it, "If you're not continuously red-teaming your AI, you're already compromised." The point is this, it's not just about security testing anymore – it's about creating a comprehensive framework of safety measures, ethical guidelines, and regulatory oversight.I immediately think of Wimpy's line from Popeye cartoons: "I'll gladly pay you Tuesday for a hamburger today," because in the end, a cheaper AI model might save you money today, but the cost to society could be much higher tomorrow. The time for voluntary compliance and assuming positive intent is over – we need robust regulations and enforcement mechanisms to ensure AI development prioritizes safety and ethics alongside innovation. NOW!

DeepSeek's Wild West: When AI Safety Takes a Back Seat

<figure id="figure-bgmpcgjgkum6m6u6zw" data-full-width=""><img id="bgmpcgjgkum6m6u6zw" src="https://cdn.durable.co/blocks/85fv06Bkm60FObkXT1rhFCV3U5Yu7AjIiifvk4ZBJNeva02J7tSvdJLxtP6erRIm.jpg" alt="" title="" class="object-cover bg-gray-100 " style="aspect-ratio: 3/2; object-position: 50% 70%;" data-corners="default" data-aspect="3:2" data-caption="" data-positionx="50%" data-positiony="70%" data-fullwidth="true" data-mce-src="https://cdn.durable.co/blocks/85fv06Bkm60FObkXT1rhFCV3U5Yu7AjIiifvk4ZBJNeva02J7tSvdJLxtP6erRIm.jpg" data-mce-style="aspect-ratio: 3/2; object-position: 50% 70%;"></figure>Thrilled to share my insights in Scott Clark's latest article: "Human-Centered AI: Building Empathy and Creativity in Machine-Driven Era"As CEO and Managing Partner at Fusion Collective, I had the opportunity to discuss a crucial aspect of our AI-driven future: the mindset that will define success in this transformative era.In my conversation with Scott, I emphasized that success in the AI age demands an abundance mindset focused on growth. According to the World Economic Forum's Future of Jobs Report, the most critical skills moving forward will center around:• AI and big data • Networks and cybersecurity • Technological literacy • Creative thinking • Resilience and adaptability • Curiosity and lifelong learningSpecifically, I told Scott, "machines are not curious by nature, and as technology advances, we all must be able to ask better questions moving forward."The article explores how human-centered AI is reshaping industries while highlighting the irreplaceable value of human skills. It's not about AI versus humans—it's about creating a future where technology amplifies our uniquely human capabilities.&nbsp;➡️ Read it <a data-mce-href="https://www.cmswire.com/customer-experience/human-centered-ai-building-empathy-and-creativity-in-machine-driven-era/" href="https://www.cmswire.com/customer-experience/human-centered-ai-building-empathy-and-creativity-in-machine-driven-era/" target="_blank" rel="noopener" data-mce-selected="inline-boundary">here.﻿</a>What skills do you think will be most crucial in the AI era?Share your thoughts and help keep this important conversation about our collective future going!

Featured In CMSWire: "Human-Centered AI: Building Empathy and Creativity in Machine-Driven Era"

<figure id="figure-3hi1ujb68tm6jnhypm" data-full-width=""><img id="3hi1ujb68tm6jnhypm" src="https://cdn.durable.co/blocks/28ghbAM40iggVp2As2ibcgL8AfdlFDQd8eGNCB9aJJodkK9BFXigQU2A0Zjw0Cus.jpg" alt="" title="" class="object-cover bg-gray-100 " style="aspect-ratio: 1/1; object-position: 100% 20%;" data-corners="default" data-aspect="1:1" data-caption="" data-positionx="100%" data-positiony="20%" data-fullwidth="true" data-mce-src="https://cdn.durable.co/blocks/28ghbAM40iggVp2As2ibcgL8AfdlFDQd8eGNCB9aJJodkK9BFXigQU2A0Zjw0Cus.jpg" data-mce-style="aspect-ratio: 1/1; object-position: 100% 20%;"></figure>Many of Steve Jobs’ quotes fit perfectly in the realm of the daily acceleration of new emerging tech that everyone wants to play with. It’s new, it’s sleek and damn is it fun to figure out! &nbsp;Hey, I’m all for playing but we are operating in a world where it’s the wild, wild west and there’s no sheriff around reinforcing rules of law because – at this point there aren’t any. NO generally agreed upon rules or guidelines around responsible use or AI ethics exist, or at least, none that are followed. We are literally assuming that the companies making this stuff are doing the right things. And, as history has shown us, that is a bad assumption any end-user should make.&nbsp;Now, DeepSeek is the new shiny IT tool to play with and I’m all for play but as with any new technology, regardless from which country or company it’s from, you need to be very careful about how you use it, what you share, and play safe. &nbsp;Here are a few reasons why you should think twice and three times before you download and start using DeepSeek. Yet another reminder, stop blindly clicking ACCEPT terms and conditions without reading them because you literally are giving them all of your data to use it, sell it and disclose it anyway they see fit. &nbsp;Yes, the AI model is at least on par with (and some would argue even more advanced than) OpenAI’s ChatGPT or Anthropic’s Claude. But there’s a serious catch.&nbsp;I’m giving it to you straight – no chaser. Make no mistake, DeepSeek is sending your data to China. And they tell you that straight up in the English-language <a data-mce-href="https://platform.deepseek.com/downloads/DeepSeek%20Privacy%20Policy.html" href="https://platform.deepseek.com/downloads/DeepSeek%20Privacy%20Policy.html" target="_blank" rel="noopener">DeepSeek privacy policy</a>, which lays out how the company handles user data; it's unequivocal: “We store the information we collect in secure servers located in the People's Republic of China.” See how reading becomes fundamental?&nbsp;In other words, all the conversations and questions you send to DeepSeek, along with the answers it generates, are being sent to China. What’s scarier is the information it collects about you. Again, this is listed plainly in DeepSeek’s privacy policy. So, what do they collect? Hang on to your hat because it’s literally everything: &nbsp;All information that you share with DeepSeek or “user input.” This is a broad category and will likely cover your chats with DeepSeek via its app or website. “We may collect your text or audio input, prompt, uploaded files, feedback, chat history, or other content that you provide to our model and Services,” the privacy policy states. However, within DeepSeek’s settings, it is “possible” to delete your chat history. Here’s the rub, while you delete your chat history – they still collected the data, sitting in China, on Chinese servers not subject to US laws.&nbsp;Think about that…even the personal data you use to setup your account goes to DeepSeek which includes your email address, phone number, date of birth, username, and more. Even if you reach out to them for questions or troubleshooting, all that data will be shared as well.It doesn’t stop there. As with all digital platforms—from websites to apps—there can also be a large amount of data that is collected automatically and silently when you use the services. DeepSeek says it will collect information about what device you are using, your operating system, IP address, and information such as crash reports. It can also record your “keystroke patterns or rhythms,” a type of data more widely collected in software built for character-based languages. Additionally, if you purchase DeepSeek’s premium services, the platform will collect that information. It also uses cookies and other tracking technology to “measure and analyze how you use our services.”So far, let’s recap what they are already telling you they’re collecting:&nbsp;<ul><li>Your DeepSeek prompt and chat history&nbsp;</li><li>Any files you upload - any files&nbsp;</li><li>Your email address</li><li>Your date of birth</li><li>Your phone number</li><li>Your device info</li><li>Your IP address </li><li>Any communications you have with DeepSeek</li><li>And the one that is potentially most concerning, keystroke patterns&nbsp;</li></ul>So, why is keystroke pattern concerning? What does that mean, keystroke patterns? DeepSeek captures your keystroke patterns so if I’m trying to mimic you, I have to mimic how you type on a computer.  That’s a significant RED FLAG privacy concern. Why do they need that and what is it going to be used for? My Spidey senses are firing!&nbsp;Let’s use a relatable and real-world example. DeepSeek captures the things that you type and typed before you even hit SUBMIT. It’s captured. How many times have you started writing your prompt in the prompt window, revised it and even sometimes just deleted the whole thing to start from scratch? In this example, if you start typing in the prompt window a question but deleted and never hit submit, it’s still captured by DeepSeek. &nbsp;And this is keystroke capture.There are a lot of inherent problems with how DeepSeek will use your data, but most concerning is even before you hit submit its stored on a server, in China, so subject to Chinese law not US law, collecting way more information you think and probably even more than that. And China has some remarkable legal requirements under its cybersecurity and privacy laws, including laws that demand tech companies cooperate with national intelligence efforts.The difference between DeepSeek vs Claude or ChatGPT is this - DeepSeek is a black hole for your data and privacy. &nbsp;Absent of international ethical use laws and frameworks, it’s figuratively and literally the wild west. Companies can, will, and are doing whatever is in their best interests.<h3>Why Should You Be Concerned?</h3>Well, most of us don’t really read the terms of service or privacy policies for a host of reasons. First, they are super long, filled with legalese and mind-numbingly boring. This is done on purpose which makes it exceptionally dangerous. So, what do we do? We scroll to the bottom and click ACCEPT. And by clicking accept you’ve literally lost control of your data and privacy. At the time this isn’t registering as a concern because you really just want to use this new shiny tool. We should’ve learned from the massive leaks and hacks that have been on the rise, but alas, we still click ACCEPT without reading. This fierce need does something that tosses concern and commonsense out the window. Maybe it’s the anticipation and excitement to use this new thing that has users blindly clicking accept but no one can afford to ignore the importance of data security.&nbsp;Case in point: DeepSeek is subject to government access under China's cybersecurity laws, which mandate that companies provide access whenever the Chinese government demands it. We don't know how many AI models are trained or how they operate, and that's concerning, too, especially if your data could be misused or maliciously exploited. And for sure you don't want your identity stolen or your bank account information in the wrong hands. Right?So, by all means play, but play safe. Remember, the more personal information you give to a chatbot or use a credit card to pay for it, and that data is stored by a company which is later hacked or improperly shared — you could be in a world of trouble. The damage done is more than being bothered because now you must change your passwords, but it could lead to you potentially losing your life's savings. Sure, companies like DeepSeek, even Meta or OpenAI, may not actively steal your information to take your identity. But still, no one is immune to cyberattacks which are on the rise — and if they're hosting your data, your data can be taken. Did you know on January 28, 2025, DeepSeek faced "<a data-mce-href="https://apnews.com/article/deepseek-ai-artificial-intelligence-be414acadbf35070d7645fe9fbd8f464" href="https://apnews.com/article/deepseek-ai-artificial-intelligence-be414acadbf35070d7645fe9fbd8f464" target="_blank" rel="noopener">large-scale malicious attacks,</a>” which forced the company to temporarily limit new registrations? We are living in a digital age that is not simple like Candyland, it’s more likened to a mashup of Game of Thrones meets Dungeons and Dragons and Warcraft.<h3>So, what can YOU do?</h3>First things first, remember, not just with DeepSeek but with any similar platform, when you use their services, you’re doing work for them, under the guise of helping you. REMEMBER, they are not working for you and your use of their platform is rigged to work for them and them alone.Second, safeguarding your data not just from DeepSeek but as well as all platforms is uber critical. Don’t be sheep and blindly provide your personal or other data wherever you go grazing from one platform to another giving them your personal life details and data. Be proactive when it comes to your digital footprint and cybersecurity. Which means, before clicking ACCEPT, actually read with scrutiny, all terms and conditions of any platform you engage with, so you clearly understand where your data is stored, how it’s being used, and who has access to it.There’s only so much the end user can do because without strong data privacy laws that apply to ALL companies, the onus to keeping your data safe and protecting your privacy falls squarely on you. All the risk is yours if things go sideways. So, manage your data and privacy like you own it – because you do.So, before you jump in the DeepSeek waters, ask yourself is my data really safe?

Stay Hungry, Stay Curious but Be Safe

<figure id="figure-nn9ezjl0gm6i6javp" data-full-width=""><img id="nn9ezjl0gm6i6javp" src="https://cdn.durable.co/blocks/1b3lmqvx5GG54VDJSkXsRfcQcOdCPzRghqsWeQcGaKzlboJssZkpyY3WovqaX7vF.png" alt="" title="" class="object-cover bg-gray-100 selected" style="aspect-ratio: 16/9; object-position: 55% 40%;" data-corners="default" data-aspect="16:9" data-caption="" data-positionx="55%" data-positiony="40%" data-fullwidth="true" data-mce-src="https://cdn.durable.co/blocks/1b3lmqvx5GG54VDJSkXsRfcQcOdCPzRghqsWeQcGaKzlboJssZkpyY3WovqaX7vF.png" data-mce-style="aspect-ratio: 16/9; object-position: 55% 40%;" data-mce-selected="1"></figure>Honored to be featured in Colin Ingram's important new article: "Does Your AI See Everyone? <a href="https://nykdaily.com/2025/01/does-your-ai-see-everyone-the-skin-detection-problem-in-self-driving-cars-and-data-analytics/" target="_blank" rel="noopener" data-mce-href="https://nykdaily.com/2025/01/does-your-ai-see-everyone-the-skin-detection-problem-in-self-driving-cars-and-data-analytics/">The Skin Detection Problem in Self-Driving Cars and Data Analytics"</a>As a technology leader passionate about inclusive innovation, I long shared my concerns about a critical issue in autonomous vehicle development: self-driving cars are struggling to detect pedestrians with darker skin tones—and the problem is worsening. And as a Black woman, this is personal not only for me but for all black and brown communities.In my conversation with Colin, I emphasized that "beneath this shiny promise of innovation is a grave worry — a possible blind spot that might have a disproportionately negative effect on the safety of Black and Brown people everywhere."The article highlights crucial findings:• New joint research from King's College London and Peking University shows alarming disparities in pedestrian detection• The impact is particularly severe in urban communities where walkability is essential• Current AI datasets predominantly represent privileged demographics, creating structural barriersI've long advocated that technology must serve everyone, regardless of gender, race, culture, or socioeconomic background. This isn't just about cars—it's about ensuring our technological future truly includes everyone.&nbsp;So, take a read of Colin's thorough investigation into why diverse datasets and inclusive testing standards aren't just nice-to-haves—they're matters of life and death.What are your thoughts on ensuring AI development represents all communities?

Featured In NYKDaily: The Skin Detection Problem in Self Driving Cars & Data Analytics

<figure id="figure-swdxpbwbhim6ecvc8b"><img id="swdxpbwbhim6ecvc8b" src="https://cdn.durable.co/blocks/8hNFrjDIrsDPAqwxzrYbQyXeDJAfUIStlAmlKN2aS9umETUL3QKnpGIsu5fpC8d8.jpg" alt="" title="" class="object-cover bg-gray-100 " style="aspect-ratio: 16/9; object-position: center center;" data-corners="default" data-aspect="16:9" data-caption="" data-positionx="undefined" data-positiony="undefined" data-fullwidth="false" data-mce-src="https://cdn.durable.co/blocks/8hNFrjDIrsDPAqwxzrYbQyXeDJAfUIStlAmlKN2aS9umETUL3QKnpGIsu5fpC8d8.jpg" data-mce-style="aspect-ratio: 16/9; object-position: center center;"></figure>The pace of innovation today is nothing short of breathtaking. For the geek in me, it’s like every day is a playground of possibilities. But let me hit the brakes for a moment and ring an alarm bell: privacy and safety. Because when an AI model can watch how you operate your computer—and even control it—this isn’t just innovation; it’s a watershed moment, and not necessarily in the best way. &nbsp;&nbsp; &nbsp; Last week, OpenAI unveiled Operator, an AI marvel that takes multitasking to a whole new level. Picture this: Operator watches your on-screen activity, processes screenshots to understand your computer’s state, and then uses its simulated keyboard and mouse to click, type, and scroll for you. Cool? Absolutely. Terrifying? Also, yes.<h3>Operator: The Friend Who Might Be Your Frenemy</h3>Think of Operator as a digital twin—a friend that mirrors your every move. It knows your quirks, habits, and routines and steps in to make your life easier. Need to order groceries? It’s got your back. Forgot to respond to that email? No problem. &nbsp;&nbsp; &nbsp; But here’s the twist: this friend isn’t just watching—it’s taking notes and reporting back to someone who may not have your best interests at heart. Every website you visit, every spreadsheet you edit, every task you complete becomes data sent to OpenAI’s cloud servers. Suddenly, your helpful friend starts feeling less like a confidant and more like a frenemy with a loose mouth. Do you really want someone—or something—that knows everything about you to also hold the power to act on your behalf? Today’s value system is warped, and today’s version of capitalism is to maximize profits. So, how can you trust a system that doesn’t value you but only values amassing more and more piles of cold hard cash? There’s a fundamental disconnect. Again, I ask, how can you blindly trust they are looking out for you?<h3>The Safety Features (and What They Don’t Solve)</h3>OpenAI insists it’s put safeguards in place: Operator requires user confirmation for sensitive tasks like sending emails or making purchases. It even restricts access to certain website categories—gambling, adult content, etc. But here’s the million-dollar question: Why are we still relying on others to define what safe looks like for our data? &nbsp;&nbsp; &nbsp; Privacy isn’t a one-size-fits-all deal. What feels secure to OpenAI’s engineers might not pass the sniff test for the rest of us. Meanwhile, regulators are crawling along at a snail’s pace, trying to keep up with tech moving at the speed of light. In 2024, we saw a new AI model debut every 2.5 days. Now, barely a month into the new year, here comes Operator—taking screenshots of your screen and acting on your behalf.<h3>“Color Me Skeptical”</h3>Operator is built on large language model transformer tech, which historically hasn’t exactly been Fort Knox. These models are notoriously vulnerable to jailbreaks and prompt injections—those clever little hacks that bypass safeguards. OpenAI says it’s implemented real-time moderation to catch these tricks, but even their internal testing missed one. One! And that’s in a controlled environment. &nbsp;&nbsp; &nbsp; <a data-mce-href="https://simonwillison.net/2025/Jan/23/introducing-operator/" href="https://simonwillison.net/2025/Jan/23/introducing-operator/" target="_blank" rel="noopener">Simon Willison</a>, a seasoned expert on AI security, summed it up best: “Color me skeptical.” If history is any guide, the dark web’s black-market masterminds are already cooking up ways to exploit this tech. Remember 2024? Cybercriminals embraced AI chatbots faster than law enforcement could say “unauthorized transaction.” Operator could easily become the next chapter in that story.<h3>Read the Fine, Fine Print</h3>OpenAI openly acknowledges the risks in its documentation: “Certain challenges and risks remain due to the difficulty of modeling the complexity of real-world scenarios and the dynamic nature of adversarial threats.” Translation: They’re trying, but they know this isn’t foolproof. &nbsp;&nbsp; &nbsp; And about privacy—let’s talk about the elephant in the room. Operator requires significant computing power, so it doesn’t run locally. Every screenshot it processes gets sent to OpenAI’s cloud servers. Every. Single. One. That’s a whole lot of trust to put in a company, whose sole purpose is to make money, no matter how many privacy controls they tout. &nbsp;&nbsp; &nbsp; Yes, OpenAI has added user-friendly features like opt-out settings, one-click data deletion, and “takeover mode” to block screenshots during sensitive tasks. Sounds great in theory, but are you really comfortable with this level of oversight? I’m not.<h3>Play But Play Smart, Stay Safe</h3>So, if you’re planning to play with Operator, let me give you some friendly advice: &nbsp;&nbsp; &nbsp; 1. Start a fresh session for each task to limit what Operator can see. &nbsp;&nbsp; &nbsp; 2. If you’re letting it spend your money, babysit the process. Get it to checkout, input payment details manually, and wipe the session afterward.&nbsp; &nbsp; 3. If anything, and I do mean anything you are working on is a trade secret, under NDA, or otherwise sensitive, do not use this tool. &nbsp;Just don’t. No reason to risk it, just let this one play out.<h3>What Keeps Me Up at Night</h3>I’m not afraid of technology. In fact, I love it. I geek out over every new gadget and groundbreaking release. What worries me is this: these models aren’t inclusive, and the ethical guidelines for their development are practically nonexistent. It doesn’t matter if your model runs locally or in the cloud. &nbsp; &nbsp; It’s one thing to invite technology into your life; it’s another to let it redecorate the place and invite its friends over. Operator may be the shiny new tool you think you’ve been waiting for, but remember, shiny doesn’t always mean safe. So, before you hand it the keys to your digital castle, ask yourself: Is the gate locked, and do you trust who’s standing outside?

Shiny, Smart, and Suspect: What OpenAI’s Operator Means for You

<figure id="figure-wqh61tzzwmm689qexu" class=""><img id="wqh61tzzwmm689qexu" src="https://cdn.durable.co/blocks/32KPhBXMhXMkh87QxBd07gzZu2MZGvy0qkwGJSf9G0F3hEn1OnYDwVs4iAZkPrKy.jpg" alt="" title="" class="object-cover bg-gray-100 " style="aspect-ratio: 3/2; object-position: 45% 35%;" data-corners="default" data-aspect="3:2" data-caption="" data-positionx="45%" data-positiony="35%" data-fullwidth="false" data-mce-src="https://cdn.durable.co/blocks/32KPhBXMhXMkh87QxBd07gzZu2MZGvy0qkwGJSf9G0F3hEn1OnYDwVs4iAZkPrKy.jpg" data-mce-style="aspect-ratio: 3/2; object-position: 45% 35%;"></figure>As consumers/end users, we can no longer afford to sit idly by and let others run roughshod over our data.It’s our data and we own it. So, why are we so willing to let go and let others use it under the guise of helping to make your life easier or to be more productive? While that is all good, because who wouldn’t want to offload mundane routine tasks so you can get more “real stuff done?” But that helpfulness and increased productivity shouldn’t require unfettered access to your data and then allow others to do whatever they see fit with it, from using it to “learn” more about your habits and “needs” to help you or selling it to 3rd parties for profit so they can sell you more “stuff.”&nbsp; &nbsp; In the advent of impending agentic AI, which is going to be here sooner than you think, as it’s reported that OpenAI in the next coming weeks will announce a next level breakthrough that unleashes PhD level super agents to do complex human tasks. On the books is a closed door <a data-mce-href="https://www.axios.com/2025/01/19/ai-superagent-openai-meta?utm_source=tldrnewsletter" href="https://www.axios.com/2025/01/19/ai-superagent-openai-meta?utm_source=tldrnewsletter" target="_blank" rel="noopener">meeting</a> scheduled with US government officials in Washington on January 30, 2025.&nbsp; &nbsp; Coupled with the deployments of new AI assistant systems like Google’s scam call protection and Apple Intelligence (refer to my previous post), both put AI EVERYWHERE on your phone – even in the middle of your private messages. Again, refer to my previous post on how to disable Apple Intelligence on your iPhone. So, have you really thought about your privacy? &nbsp; &nbsp; The looming question continues to be this - what are the privacy implications of AI? This keeps me up at night especially due to the recent EU debate over <a data-mce-href="https://www.patrick-breyer.de/en/posts/chat-control/" href="https://www.patrick-breyer.de/en/posts/chat-control/" target="_blank" rel="noopener">mandatory content scanning laws</a> that would require machine learning systems to scan virtually every private message you send. &nbsp;So, what happened to end-to-end encryption and better yet, what does it mean for it in the future?&nbsp; <h3>A 101 Primer on End-to-End Encryption</h3>For the past 10 years or so, it was all the rage as we saw the explosion of end-to-end encrypted communication platforms. Remember prior to 2011, most cloud connected devices uploaded their data in plaintext, and you know the hot mess that ensued from there. For many, their private data was exposed like laundry out to dry on the line for everyone from hackers, subpoenas, warrants or exploited by the platforms themselves. New data storage approaches started to evolve around 2011, and we saw messaging apps like WhatsApp and Apple’s iMessage rolling out default end-to-end encryption for private messaging.How does this work?Simply put, this technology changed key management to ensure that a server would never see the plaintext of your messages. Cool, right? Not too long afterwards, phone designers starting encrypting data stored locally on your phone. The recurring theme for both advancements? Nothing needed to be processed by a server. &nbsp; So, here’s the rub: For data that requires intense processing options are limited. Sure, for cloud backups and private messages it’s easy to hide content from servers. But when you want to do serious compute, the typical phone has a limited amount of processing power, including limited amounts of RAM as well as battery capacity. Case in point, did you see how your battery life significantly took a hit if you didn’t disable Apple Intelligence? Even if you wanted to do something like photo processing on the most tricked out iPhone, it still needs to be plugged in to avoid draining the battery. Not to mention, there’s a cornucopia of variation in phone hardware. There are some tricked out phones with onboard GPUs and neural engines, but they aren’t cheap. Most start around $1,400 and go up from there. Not only are these phones not typical, but they are also cost prohibitive and what’s left will range in their processing capability.&nbsp; &nbsp; Now with the primer on the evolution of end-to-end encryption, how does AI come into play? If you weren’t paying attention because the cat memes got you on Instagram, in 2024 there was a new frontier model released every 2.5 days. These new powerful models have cutting edge capabilities ranging from LLMs that can generate and understand human text to new image processing models that can produce content on their own from text, images, audio, code, 3D and more.We are entering unchartered waters, and no one really knows where this can go or actually knows how useful these models can be to their customers but that hasn’t stopped any and every tech company to find applications for this tech – including phone and messaging companies. The collective rally cry and race to the bottom is – models are the future. Personally, I believe the onslaught of integrating AI into our daily lives via our mobile phones is likened to the amuse bouche. I’m a self-proclaimed foodie so, let me explain. Have you ever been to a restaurant and the chef sends out something called the “amuse bouche?” It’s a French term for 'mouth amuser' and chefs serve this to diners to enjoy while waiting for their food orders to arrive as well as a way of saying 'welcome' to guests and pampering them with something special. &nbsp;Well, consider this the amuse bouche for the main entrée to come – the deployment of AI Agents. &nbsp; &nbsp; Theoretically, with these new agents you don’t need to bother reading and responding to messages on your phone because they’ll read your email and text messages and answer them for you. They can order your food, find you the best deals on shopping – basically anticipate your every want or need. And how can they do this? Unfettered unrestricted access to all your private data coupled with a considerable amount of compute power to process it.&nbsp; &nbsp; So, let’s go back to the end-to-end encryption and data storage primer. The key here is that nothing needed to be processed by a server. But now, these agents will require more compute power than most phones have so they can run these powerful models. And that means, much of the processing as well as the data that needs to be processed will have to be offloaded to remote servers.&nbsp; &nbsp; Huh? Yep, so what happens to all that privacy?&nbsp; &nbsp; This is why I implore everyone to be very careful before enabling all of this cool tech to do cool stuff to that makes your life easier, because the privacy implications are enormous. All of this slickness promised requires compute that will be done off your device which means sending considerably more of your private data to systems that will examine and summarize your data, produce compact summaries of your LIFE. These systems will eventually know everything about you and your friends, because they will have access to our most intimate and private conversations. We are standing at the precipice of the complete death of privacy and now is the time to ask the hard questions about these systems. Especially existential questions like: are they really working for us?&nbsp; &nbsp;&nbsp; When it comes to end-to-end encryption, current messaging systems issue very specific technical guarantees. By definition, an end-to-end encrypted system is designed to ensure that plaintext message content in transit is not available anywhere except for the end-devices of the participants and anyone the participants or their devices choose to share it with. I want to lean in on this because this is super important. If you are 1:1 texting cool but what if you start a WhatsApp group chat and someone else in the group decides to turn on some service that uploads (your) received plaintext messages to WhatsApp. Can you say the message you received accurately describes how WhatsApp treats your data? &nbsp;<figure id="figure-xyhl9qa9d9m688zs1g" class=""><img id="xyhl9qa9d9m688zs1g" src="https://cdn.durable.co/blocks/16GNnayE13BZF9zEUnywmZUgE5bRoPlrO5n3UPmdGds9pSaDiehM14lRH60iO6Za.png" alt="" title="" class="object-cover bg-gray-100 " style="aspect-ratio: 16/9; object-position: undefined undefined;" data-corners="default" data-aspect="16:9" data-caption="" data-positionx="undefined" data-positiony="undefined" data-fullwidth="false" data-mce-src="https://cdn.durable.co/blocks/16GNnayE13BZF9zEUnywmZUgE5bRoPlrO5n3UPmdGds9pSaDiehM14lRH60iO6Za.png" data-mce-style="aspect-ratio: 16/9; object-position: undefined undefined;"></figure>This is where things get even more complicated because now we are talking about informed consent. Consent is both a moral and legal concept and the law is different in every corner of the world. &nbsp; &nbsp; Getting back on my soapbox about being intentional about owning your data and protecting your privacy, some companies will do an admirable job at letting their users know exactly what’s happening as a way to earn trust. Others, well, will bury your consent inside of an inscrutable lengthy riddled with legalese “terms of service” document you never read but click “accept” so you can go about your business. In the EU they’ll probably just invent a new type of cookie banner.&nbsp; &nbsp; We are on an accelerated path to a world where you should expect more and more of your data to be processed by AI and that processing will most likely be done off your device. Altruistic end-to-end encryption services will actively inform you when this happens which presents you an option to opt out. But, if it’s everywhere then our options will probably be limited. &nbsp; &nbsp; I’ve said this and will continue to say it. I am not afraid of technology, I actually geek out over all the new emerging technology coming out. But what worries me and keeps me up at night? These models are not inclusive, they are woefully absent of regulations on ethical development, deployment, and use. It doesn’t really matter what technical choices we make around privacy. It does not matter if your model is running locally, or if it uses trusted cloud hardware — once a sufficiently powerful general-purpose agent has been deployed on your phone, the only question that remains is who is given access to talk to it. &nbsp; &nbsp; And the question you should think about is - will it only be you?

Privacy Implications of AI Agents & Data Processing

Yvette Schmitter

My name is Yvette Schmitter, and I’m a dynamic, international motivational, keynote speaker and technology leader on a mission to transform the future of tech into a force that includes and uplifts us all. Drawing from my expertise in AI, cloud computing, and digital transformation, I empower audiences to lead with courage, dismantle barriers, and create solutions that bridge divides and drive lasting impact. Let’s build a future where innovation isn’t just progress—it’s progress for all. Join me in shaping the future we all deserve. My mission is to redefine the future of technology by championing innovation that includes and empowers all. As a trailblazer in the tech industry, I lead with purpose and passion, ensuring that every room where decisions are made is filled with diverse perspectives, so that technology truly serves humanity—not just a privileged few. By breaking down barriers and building bridges, I’m committed to creating a tech landscape where every voice is heard, every experience is valued, and every community has the opportunity to thrive.


Tech & Tell

DeepSeek's Wild West: When AI Safety Takes a Back Seat

Featured In CMSWire: "Human-Centered AI: Building Empathy...

Stay Hungry, Stay Curious but Be Safe

Featured In NYKDaily: The Skin Detection Problem in Self ...

Shiny, Smart, and Suspect: What OpenAI’s Operator Means f...

Privacy Implications of AI Agents & Data Processing