Combating Disinformation

Cyber Smart from the Start: Defending Finland’s Future in the Classroom Finland has long been celebrated for its world-class education system and commitment to digital innovation. But as technology becomes increasingly entwined with everyday life, new challenges are emerging—especially for the next generation. The rise of misinformation, cyberbullying, and online fraud means that teaching traditional subjects is no longer enough. Today’s students must be equipped with the tools to think critically, act safely, and defend themselves in the digital world. Disinformation campaigns, particularly from hostile foreign actors like Russia, have become more frequent and more sophisticated. These campaigns are not limited to military or political targets—they affect everyday citizens, manipulating emotions, distorting facts, and undermining democratic values. Finnish students must be taught how to recognize propaganda, question suspicious sources, and resist the temptation to share unverified information. But media literacy alone won’t cut it. Our young people also need to understand personal cybersecurity—from using secure passwords and avoiding phishing scams, to managing their online identity and digital footprint. By integrating cybersecurity and disinformation awareness into the national curriculum, we can ensure that Finnish students grow up not just smart, but cyber smart—ready to protect themselves, and their country, from the digital threats of today and tomorrow. #cybersecurity #education #Finland #CyberHygiene #misinformation #disinformation #PrimarySchool #SecondarySchool #privacy #WhyCantWeDoThatHere #democracy

I enjoyed being interviewed by Brian O'Donovan of RTÉ news for this story (link below) on that crazy deepfake that was released with just a couple of days left in the Irish Presidential election. The deepfake video of Catherine Connolly announcing her withdrawal from the presidential race illustrates how powerful and accessible generative AI has become. It convincingly replicated both Ms Connolly’s appearance and voice, wrapped within a fabricated RTÉ News broadcast. Under the EU’s Artificial Intelligence Act, this kind of material is no longer viewed as a novelty. It is a regulated phenomenon, formally recognised as a “deepfake”. Article 3(60) defines a deepfake as AI-generated or manipulated image, audio, or video content that resembles a real person, object, place, or event and would falsely appear authentic to the viewer. In essence, it is synthetic content that convincingly imitates reality. The AI Act deals with deepfakes under Article 50, which sets out transparency obligations for both developers and users of such systems. Providers of AI models that generate images, video, or audio must build in technical mechanisms, such as watermarks, cryptographic signatures, or metadata - that identify outputs as artificially generated. These identifiers must be robust, interoperable, and detectable in machine-readable form. But that doesn’t stop users from screenshotting and cropping - which removes these signatures. Deployers, meaning those who use these systems to publish or distribute content, have a separate duty to disclose when material has been created or manipulated by AI. The disclosure must be clear and easily recognisable. There are narrow exceptions, such as for artistic, satirical, or law enforcement purposes, but even then some form of notice is generally required. What strikes me as interesting is - could the person/country who generated the deepfake video claim it was satire? The Facebook page the video came from had a notice that it was AI generated (which went unnoticed by most). That’s troubling. Failure to label or mark synthetic content can, in certain circumstances, trigger the AI Act’s prohibition on manipulative practices under Article 5(1)(a). If the omission materially distorts a person’s behaviour or decision-making, particularly in a way that could cause significant harm, it may fall within the category of banned AI practices. The Connolly incident shows why these provisions matter. Without visible disclosure, a deepfake risks undermining electoral integrity, public confidence, and informed democratic participation. The EU’s framework does not outlaw deepfakes outright, but it insists on transparency and accountability in their creation and use. As models capable of hyper-realistic synthesis continue to evolve, the legal focus is shifting from censorship to traceability - ensuring that authenticity can be verified even when the human eye can no longer tell the difference.

Community-based fact-checking is a promising approach to correct misleading posts. Exposing users to community notes reduces the subsequent spread of misleading posts by 61.2% (an increase the odds that users delete their misleading posts by 94.3%) Although community notes are effective in reducing the spread of posts once annotated, they usually appear too late to intervene in the early and most viral stage of the diffusion. As a result, their system-wide effect is modest. This is based on an analysis of 237,180 community fact-checked cascades reposted more than 431 million times https://lnkd.in/eWNRchek One reason I think they are effective is because they harness social norms--if other people *like me* find the comment misleading that is often more compelling that a simple fact check. We found that a more effective strategy might be adding a "misleading" button right on social media posts. This allows people to see social norms much faster (before content goes viral). This is 5X more effective than other strategies for fighting misinformation and does not require fact-checkers or official content moderation. In additional, it works really well for polarizing content--which community notes struggle to stop. https://lnkd.in/eAAVg6mP

Trust in American institutions has been declining for years. Today less than half of Americans trust health care leaders, and health care journalism is rated last in terms of trust from America’s public—all according to the 2024 Edelman Trust Barometer. While some researchers say the phenomenon of mistrust isn't new and has come in waves across a century of American history, the recent Edelman findings feel especially troubling now as we look ahead to the future of US health care.    No one has all the answers on where to go from here, but as I consider the road ahead, I’m grounded in part by the strategies shared by David Rousseau and Noam Levey on separate past episodes of the podcast I host with Don Berwick, Turn on the Lights. The strategies they each offered for building public trust in journalism can be applied by health care leaders in the work we do every day --   1. Be transparent about your methods. Show people the data, sources, best practices that inform your thinking.   2. Have the humility to know you don’t always know the answers.    3. Bring in local expert voices that your community/audience connects with and trusts, and make sure those voices are diverse.   4. Use plain language, never jargon. Connect with people in their terms and on their terms.    5. Make people/patients the focus, always. Put their experiences, needs, assumptions, point of view at the center of EVERY cause, case, and communication you make.    Trust is crucial for optimal functioning of the health care system. Whether you’re a health care journalist, leader, or provider you can put these strategies to work and contribute to our collective rebuilding of trust in health care.    For more, listen to past episodes of Turn on the Lights here: https://bit.ly/3YWXL5f and explore IHI’s theory of how to repair, build, and strengthen organizational trustworthiness in health care: https://bit.ly/40MNQkh. 

🧵 New report just dropped 🚨 "Fractured Reality: How Democracy Can Win the Global Struggle Over the Information Space" — from the EU Joint Research Centre, led by Mario Scharfbillig and I. A landmark read for anyone working on disinformation, platforms & democracy. 👇 1/10 The core problem: the attention economy. Platforms maximise engagement, not truth. Algorithms reward negative, emotional & divisive content — not because of malice, but because that's what keeps eyes on screen. Democracy pays the price. 2/10 Disinformation has evolved. We've moved beyond isolated false claims to a "Fantasy-Industrial Complex" — a loosely coordinated ecosystem that floods the zone with distrust, distraction & anti-democratic sentiment. The goal isn't to convince — it's to confuse. 3/10 Key stat: misinformation exposure hovers around 1–10% of civic content online, rising to 10–30% for contentious topics \(climate, health, Russia/Ukraine\). And 47% of older adults hold misinformed beliefs about health issues. Scale matters. 4/10 Generative AI is amplifying this further. While evidence it's been a "game changer" in elections is still limited, it creates the illusion of knowledge — especially when blended with social media. The threat is real, even if still emerging. 5/10 The report's first fix: don't blame users. Blaming individuals for being misinformed is a misdiagnosis. We need systemic reform — better platform design, crowdsourced knowledge \(like Wikipedia\), media literacy, and fact-checking that actually reaches people. 6/10 Second fix: reform the business model. Proposals include a progressive digital advertising tax, interoperability & in-situ data rights so users can choose their own algorithms. As long as misinformation is profitable, it will persist. 7/10 Third fix: EU digital sovereignty. Most major platforms are foreign-owned & may not share European democratic values. The report backs decentralised alternatives — think Mastodon, FediVerse, Eurosky — and calls for a "European CERN for Data & Democracy." 8/10 77% of experts surveyed call the current trajectory — a "struggle for information supremacy" — harmful. Only 3% see it as beneficial. The solutions exist. The political will is what's needed. 9/10 Launched today at the Cambridge Disinformation Summit. Read the full report 👇 https://sks.to/fractured @scharfbillig.bsky.social @sachaltay.bsky.social @https://lnkd.in/eihFADYN @sebavalenz.bsky.social @ulliecker.bsky.social @michaelemann.bsky.social @naomioreskes.bsky.social 10/10

The temperature of social media platforms right now feels “too hot”. The real world isn’t this extreme. Social platforms thrive on discourse. “The algorithm” fuels rage-bait content that is borderline acceptable, and keeps you scrolling. 🧠 If you rely on social media as your main news source, you are setting yourself up for failure. It is easier to process and temperature-check the state of the world if you have a diverse media diet. Algorithms are designed either to feed you agreeable content that placates you, or to trigger you emotionally. Rarely do they serve up real substance. In the United States, according to the Pew Research Center, 39% of Americans under 30 say they regularly get news from TikTok. (There types of stats concern me, and what counts as “news” is often blurry. It might be breaking stories, political commentary, influencer takes or mainstream outlets repackaged in social feeds). If you spend a few minutes scrolling for your news. you could be forgiven for thinking the battle lines are drawn on partisan lines and the doomsday clock is seconds to midnight. But if you switch to respected publications with experienced journalists and analysts, the coverage is more measured, less heated and more contemplative. Credible voices steer the discourse into safer waters. And if you are deliberate, you can balance your reading across left-leaning and right-leaning outlets, to challenge your own views and develop some second-order thinking. My take: Social media and tech platforms have their place, but they are primarily entertainment platforms, not impartial news ones. That is why a diverse media diet matters. ❌ Do not get all your news from social media. ❌ Do not interpret the tone of a debate solely through what you see on social media. You are seeing the extremes. That is the algorithm at work, not a reflection of the whole conversation. ✅ Do read widely beyond social media and encourage others to do the same, especially if they are bogged down in online arguments. ✅ Do seek out traditional publications, podcasts and writers with measured perspectives. Balance left-leaning and right-leaning viewpoints if you want to build real depth in your understanding. ✅ Do use other social platforms where the makeup of users and content is different. For example, Substack often favours depth, while LinkedIn elevates expertise and promotes content made by subject matter experts and thought leaders. Is social media really a reflection of what is happening in the world, or of how most people feel? Also, if you have recommendations for strong podcasts, opinion writers or fresh publications in geopolitics or technology, I would love to hear them. #MediaLiteracy #SocialMediaTrends #NewsConsumption #AttentionEconomy Image: ChatGPT

The Guardian's recent piece on the information crisis names something most of us feel but struggle to articulate. We have moved from misinformation to deliberate disinformation, into something far more disorienting: a world where the shared foundations for evaluating truth — journalism, science, institutions — are themselves distrusted. AI-generated content at an industrial scale, deepfakes, algorithmic amplification, epistemic fragmentation, and sheer exhaustion have combined to produce what philosophers call an “epistemic crisis”. The real casualty isn't just truth, it's our capacity to agree on a shared reality. And without that, dialogue, institutional trust, and collective action all erode. In “What the Heck Do I Do with My Life”, I wrote about developing your own inner GPS to navigate chaos. That GPS is your character — the moral compass that tells you what's true, what matters, and how to act when the signals around you are scrambled. Here's what building it looks like in practice: 1. Develop critical thinking and discernment  Don't outsource your judgment to algorithms, influencers, or even experts. Ask: What's the evidence? Who benefits from me believing this? What am I not seeing? Discernment is a muscle; it weakens if you don't use it. 2. Curate reliable voices, not just reliable facts  Identify a small set of thoughtful, intellectually honest people worth following across disciplines and geographies. Not people who always agree with you, but people whose reasoning you have learned to trust. This is a long-term investment, not a quick fix. 3. Actively seek out dissenting opinions.  The algorithm will never do this for you. It's designed to do the opposite. Make it a practice to understand the strongest version of views you disagree with. You may not change your mind, but you will hold your conclusions more honestly. 4. Slow down before you believe or share  The platforms are engineered for speed and outrage. Pause and ask: Do I actually know this is true, or does it just feel true? Most viral misinformation falls apart in sixty seconds of basic scrutiny. 5. Build your character — that's your real GPS  Information literacy helps. But what ultimately navigates you through chaos is who you are: your values, your integrity, your willingness to sit with uncertainty rather than grab the nearest certainty. Character is built slowly, through choices, reflection, and honest self-examination. There are no shortcuts, but there is no better investment. The information crisis won't be solved by better fact-checkers alone. It will be navigated by people who have done the harder work of developing themselves. What would you add to this list?

AI for Crisis, Part 4: The FDA didn’t say it. But AI made it look like they did. A mid-sized biopharma company is days away from FDA feedback on a late-stage therapy. Then a PDF starts circulating all over social It looks official. FDA letterhead. Regulatory language that matches past communications. References to real personnel and guidance documents. A denial of approval. It spreads quickly. Investor forums. Telegram groups. A few biotech Substacks. Analysts repost it with light caveats (we're looking into this) Traders react anyway. The company’s stock drops 27% by noon. The real FDA response arrives three days later and it’s positive. But the market has moved on. The damage is done. This isn’t science fiction. It’s what AI-enabled forgery looks like in healthcare. No breach. No leaker. Just one fake AI generated document that's well-timed, well-crafted, and hard to disprove at speed. If you're in comms, regulatory affairs, or investor relations, here’s how to prepare: 1. Publish your trusted communication policies and protocol now: Make it crystal clear that any FDA or regulatory updates will come only through: • SEC filings • Company website or official hosted investor calls • Scheduled in-person announcements and investor day presentations • Formal press releases The order matters. Get it on your IR page and in your newsroom. Repeat it until it’s boring. 2. Build strong regulatory relationships before you need them: If a fake FDA letter drops, you need people at the agency who can verify it within minutes and help you to fix. This isn’t just coordination. It’s a reputational safety net. It should be treated as critical infrastructure in the AI era. 3. Monitor where the lies start: If you’re only watching Linkedin, Reddit and X, you’re already way behind. Track Discords, WhatsApp and Signal groups, Substacks, and investor chatrooms. This is where early disinfo testing happens and where you need eyes. 4. Prep for a “false but believable” leak: Most companies have a playbook for real bad news. You need one for fake bad news too especially when it looks real on a phone screen. Legal, regulatory, and comms all need to rehearse this scenario together. 5. Run the dang drills: Not next quarter. Now. If a fake FDA denial landed in your inbox tomorrow, would your team know what to do by hour one? If not, you’re not ready. The scariest thing about the new reality is you don’t need a real scandal to lose trust. Just a forgery that lands first. What fake stories have you had to deal with in the past couple of years?

Navigating the Maze of Truth In a world where alternative facts often muddy the waters of truth, how can individuals and communities stay resilient and well-prepared for the challenges they face? As misinformation spreads and reliable news sources are sometimes overshadowed by sensationalism, maintaining resilience becomes not just a matter of personal vigilance but a collective responsibility. 1. Develop Critical Thinking Skills - In an era where alternative facts can easily infiltrate public discourse, honing critical thinking skills is essential. This means questioning the credibility of sources, cross-referencing information with multiple reputable outlets, and analyzing the motives behind the information being presented. Encourage a mindset that doesn’t just accept information at face value but actively seeks verification. 2. Prioritize Media Literacy - Educate yourself and others about media literacy. Understanding how media works, recognizing bias, and differentiating between fact-based journalism and opinion pieces are crucial skills. Schools, workplaces, and community organizations should incorporate media literacy training to empower individuals to navigate complex information landscapes effectively. 3. Seek Out Reliable Sources - Identify and rely on reputable news sources that adhere to strict journalistic standards. These sources prioritize accuracy, fairness, and transparency. Create a diverse media diet by following established institutions and experts in relevant fields. Avoid echo chambers by engaging with a variety of perspectives while focusing on sources known for their reliability. 4. Verify Information Before Sharing - Before sharing information on social media or within your community, take a moment to verify its authenticity. Use fact-checking websites and tools to confirm details and debunk myths. By being a responsible consumer and sharer of information, you contribute to reducing the spread of misinformation and helping others stay informed. 5. Advocate for Transparency and Accountability - Support and advocate for transparency and accountability in media and information sources. Encourage media outlets and public figures to adhere to high standards of accuracy and honesty. Demand accountability from those who spread misinformation and promote a culture of truthfulness in public discourse. Our collective resilience depends on our ability to discern truth from fiction. Equip yourself with the tools to critically evaluate information and promote media literacy in your community. Stay vigilant, stay informed, and together, let’s build a more resilient and informed society. #StayResilient #CriticalThinking #MediaLiteracy #VerifyBeforeYouShare #TruthMatters

Do California’s legislative efforts to regulate AI reflect a growing concern for digital ethics, personal rights, and democratic integrity, or are they legislative overreach that will stifle innovation? There are five key bills ready to be signed into law that address issues ranging from the use of digital replicas in contracts and posthumous rights of deceased personalities to the transparency and safety of AI platforms. I’ve read them. You should, too. The links and short descriptions are below. AB 2602: Contracts: digital replicas – This bill mandates that contracts for personal or professional services involving digital replicas must clearly specify the intended uses of the replica. It also requires that individuals involved have access to legal counsel or labor union representation during contract negotiations in order to protect performers’ rights in the digital age. SB 1047: Safe and Secure Innovation for Frontier Artificial Intelligence Models Act – This bill establishes safety regulations for “covered AI models,” defined by computational power and training costs. Developers of these models must implement safety measures, conduct regular audits, and report significant incidents to the California Department of Technology. AB 2013: Artificial intelligence: transparency – This bill requires businesses that use generative AI systems to disclose any use of copyrighted materials in the training data. It also mandates that clear information about the AI system’s capabilities and limitations be provided to users. AB 1836: Deceased personalities: digital replicas – This bill prohibits the use of digital replicas of deceased personalities in audiovisual works without prior consent from their estate. It extends protections to ensure that digital replicas are not used posthumously without authorization, addressing concerns of exploitation. AB 2655: Defending Democracy from Deepfake Deception Act of 2024 – This bill requires clear disclosure of AI-generated content in political advertisements and campaign materials. It also prohibits the distribution of deceptive audio or visual media of candidates that could mislead voters, aiming to protect electoral integrity.