N1HRV — broadcast
20260808 013000 UTC
295 transcript segments
Google Cloud Speech-to-Text API (Chirp) + Gemini 2.5 Flash Non-Thinking
Data courtesy of The GDELT Project (https://www.gdeltproject.org/), from the Internet Archive TV News Archive.
Machine transcription. Treat it as a searchable index of what was broadcast, not a verbatim quotation record.

[00:00:05] about wars and other crises can put you in
[00:03:40] Cybercriminals launch attacks on power grids, hospitals, and
[00:03:43] banks. AI can protect us, but it's a constant
[00:03:47] race between hackers and developers. The
[00:03:50] latest AI model from US company Anthropic can detect
[00:03:54] gaps in systems, but in the wrong hands could also be used
[00:03:58] as a cyber weapon. Anthropic itself is warning of
[00:04:01] potential misuse out of genuine concern or is it a
[00:04:05] clever marketing
[00:04:06] strategy?
[00:04:09] This and much more on DW Science Show. Welcome to
[00:04:13] Tomorrow
[00:04:14] Today.
[00:04:19] Our digital world is vulnerable to abuse with consequences that can
[00:04:23] be devastating. US AI Company Anthropic says
[00:04:27] it has developed software that in the wrong hands
[00:04:30] can exacerbate the dangers. At first glance, it seems
[00:04:33] positive. The program has identified at least a thousand
[00:04:37] previously unknown security vulnerabilities in operating
[00:04:40] systems and browsers. I've found more bugs in the last
[00:04:44] couple of weeks than I found in the rest of my life combined. The concern
[00:04:47] making the headlines is that the program doesn't only detect flaws but
[00:04:51] can also exploit them to infiltrate systems.
[00:04:54] Anthropic says the potential threat of misuse is so high that just
[00:04:58] a few major US companies are to be granted
[00:05:02] access. The fear is that in the wrong hands,
[00:05:05] the technology could be used to attack critical infrastructure
[00:05:09] like power grids or hospitals, an alarming
[00:05:12] prospect, say many in the US
[00:05:14] tech community.
[00:05:18] I think this is going to be a kind of forced
[00:05:22] reset for the entire cybersecurity
[00:05:25] industry and a very significant
[00:05:28] event in the history of technology. So
[00:05:32] significant and so potentially harmful that the
[00:05:35] US Treasury Secretary and Federal Reserve Chair are
[00:05:38] reported to have recently held an urgent meeting with American bank
[00:05:42] executives.
[00:05:45] But is the new software really such a big threat?
[00:05:49] We talked to digital technology specialist Guido
[00:05:52] Berger.
[00:05:56] Anthropic likes to hype itself up with these apocalyptic descriptions of
[00:06:00] its models' capabilities, practically rubbing it in our
[00:06:03] faces. It's partly marketing, but it's also
[00:06:06] a real problem, and it's a problem that goes beyond
[00:06:09] Anthropic.
[00:06:20] AI is evolving at breakneck speed, and it's not uncommon for companies
[00:06:24] to address the dangers posed by developments. Sam
[00:06:28] Altman, CEO of Open AI, has warned about his own language model
[00:06:31] chat GPT, calling for society to be protected against the
[00:06:35] influence of major AI
[00:06:36] firms.
[00:06:40] AI expert Tilo Stadelmann says: AI
[00:06:43] companies launching powerful models while simultaneously
[00:06:46] warning against them is a
[00:06:48] tactic. It's also a way to
[00:06:52] emphasize your importance to
[00:06:55] investors and customers, saying we're so relevant, it's
[00:06:59] actually dangerous. You shouldn't trust anyone else with
[00:07:02] this. Also, you shouldn't just release it. Open
[00:07:06] source is very dangerous, and we need to keep the competition at
[00:07:09] bay. Those are the words of someone focused on.
[00:08:13] Hype or genuine threat? Whatever the answer, security
[00:08:16] officials and agencies are certain to face growing challenges
[00:08:19] in the
[00:08:20] future.
[00:08:24] The aviation industry is also at increasing risk from cyber
[00:08:27] threats. Jamming and spoofing attacks destabilize the
[00:08:31] plane's GPS signals, either blocking them completely or
[00:08:35] overriding them with fake signals. The aircraft
[00:08:38] can then lose its bearings and veer off course. Pilots need
[00:08:42] to be well-trained to deal with these situations in cockpit
[00:08:45] simulators, like at Wizz Air's training center in
[00:08:47] Hungary.
[00:08:56] This is what happens when the GPS signals on commercial aircraft are
[00:08:59] blocked. There's a short beep and the pilots receive a warning
[00:09:03] message. Wizz Air is a low-cost airline
[00:09:06] based in Budapest. Due to its proximity to Ukraine, the
[00:09:10] Hungarian company has become expert in dealing with jamming and
[00:09:14] spoofing, the disruption of satellite signals, an issue that
[00:09:17] affects all airlines.
[00:09:20] Wizz Air has its own state-of-the-art training center for
[00:09:23] both pilots and cabin crews.
[00:09:27] We were given unprecedented access to film simulated
[00:09:30] cyber attacks in the cockpit. There are more than
[00:09:33] a hundred potential problems that pilots can
[00:09:36] face. Today, Szabolcs Nagy is in
[00:09:39] charge of putting them through their paces.
[00:09:45] Uh, for the GPS spoofing and jamming part as well, we can
[00:09:49] we can create some scenarios, which one
[00:09:52] is for the for the navigation, for the uh, GPS, uh
[00:09:56] primary loss. We can do, we can do map shift as
[00:09:59] well, which one, in a practical uh, for the for the pilots it
[00:10:03] will be useful, if that one is happening with them in the real life, they know how
[00:10:07] to act, how to do the proper step, how
[00:10:10] to to build up some, um,
[00:10:15] Today, Ákos Pethő is flying us from Budapest to Cyprus
[00:10:19] in an A321. After takeoff he's confronted with
[00:10:22] a series of attempted interferences.
[00:10:25] Jamming is when external signals disrupt signals from the
[00:10:29] plane's GPS, which subsequently cuts out. With
[00:10:33] spoofing, false navigation information is sent to the plane's
[00:10:36] receiver, overriding the actual GPS signals.
[00:10:40] The plane is misdirected.
[00:10:43] We can now see Cyprus. The approach to Larnaca is regularly
[00:10:47] disrupted by signals from the Middle East.
[00:10:51] So, as you can see, really...
[00:10:55] slightly that our map is moving
[00:10:58] towards the left, and that's why our aircraft thinks that we are
[00:11:02] right of the clear routing. And you see, we are turning shallow it to
[00:11:05] the left, even though we've established before that we should, that
[00:11:09] we're supposed to be flying straight ahead. That's a
[00:11:12] typical
[00:11:15] uh, symptom of spoofing. This is what we call map shift. This is
[00:11:18] when the aircraft's position is uh, altered to
[00:11:22] such an extent that the auto-flight system itself
[00:11:26] doesn't really know where to fly. One
[00:11:29] solution is to turn off the GPS system, leaving the
[00:11:32] aircraft to rely on motion sensors and signals from radio
[00:11:36] beacons to navigate, like in the old days. In most
[00:11:40] cases, pilots can even react in advance thanks to very precise
[00:11:44] maps of the regions where GPS interference occurs
[00:11:46] and regular training, of
[00:11:49] course.
[00:11:51] So first of all, what we are doing to our pilots is training.
[00:11:56] And we also try to educate them, we are
[00:11:59] talking to them, we are sending them materials with which they first of all
[00:12:03] can understand the background and after that understand
[00:12:07] why it is very necessary to apply the given
[00:12:10] procedures and policies we have to
[00:12:13] avoid the worst outcomes of GPS interference to
[00:12:17] happen. Pilots have to constantly compare
[00:12:20] data from the cockpit's different navigation systems.
[00:12:24] So does all this?
[00:12:55] 1980s. Of course, spoofing was not an issue. So far there
[00:12:59] are no technical countermeasures, but one could be
[00:13:02] the European satellite navigation system Galileo. It'll be
[00:13:06] built into cockpits in the future. However, even its
[00:13:09] signals are vulnerable, which is why ESA conducted a
[00:13:12] large-scale test in Norway in which it intentionally disrupted
[00:13:16] navigation signals. ESA plans to make Galileo
[00:13:20] stronger and more resilient with the help of new low
[00:13:24] altitude,
[00:13:55] Those frequencies are monitored at facilities including the German
[00:13:58] Aerospace Center. A mobile system scans the
[00:14:01] frequency band looking for disruptions. It's something that could be used at any
[00:14:05] airport.
[00:14:11] than jamming signals. They're harder to detect. If someone were to try
[00:14:14] to actively interfere, they would transmit in these two
[00:14:17] bands.
[00:14:20] This is what a device sending false signals looks like, but it can
[00:14:23] also be smaller or much larger, where spoofing is used
[00:14:27] as a weapon. And here's a possible defense strategy also
[00:14:31] developed by the German Aerospace Center, an antenna that
[00:14:34] suppresses the jamming signals.
[00:14:37] It's designed so that it can be integrated into all current
[00:14:41] systems.
[00:14:44] These systems protect against jamming at a physical level. They make use of the
[00:14:48] fact that useful signals and jamming signals come from different
[00:14:51] directions, and that's very hard to replicate by anyone wanting
[00:14:58] That's because actual satellite signals come from above,
[00:15:01] shown here in red. That clarity allows pilots to
[00:15:05] see and work with the real GPS signals.
[00:15:08] Theoretically, the antenna is ready to go, but safety regulations
[00:15:12] mean approval processes can take years. So it will
[00:15:16] be some time before Ákos Pethő and his colleagues can
[00:15:19] land without interference all over the world again. Then
[00:15:23] passengers can enjoy. will be even safer than
[00:15:27] ever.
[00:15:31] Stories like that one can give the impression that the world is going from
[00:15:34] bad to worse, partly because we tend to focus more on
[00:15:37] negative than the positive. Consuming a steady dose of news
[00:15:43] a spiral of negativity. And yet, a lot of
[00:15:47] things have improved in recent
[00:15:48] decades.
[00:15:54] What percentage of girls worldwide attend primary school?
[00:15:57] 50%, 70% or more than
[00:16:00] 90%? We asked around.
[00:16:04] 50% at most. I'd say more like 50%.
[00:17:07] the expert, why is there such a huge gap
[00:17:11] between publicly available facts and our perception of
[00:17:15] reality?
[00:17:19] There is a term in
[00:17:21] science called negativity bias.
[00:17:25] It refers to people being more likely to look for negative information
[00:17:29] and also remember negative information. and partly also
[00:17:32] to remember negative information more
[00:17:34] strongly.
[00:17:37] Why is that? Our minds work like a filter, and this
[00:17:41] filter favors the negative.
[00:17:44] It's a legacy from evolution. After
[00:17:47] all, a hundred thousand years ago it didn't matter if someone overlooked a
[00:17:51] beautiful flower, but missing the tracks of a
[00:17:54] predator could be
[00:17:56] fatal.
[00:17:59] That's why even today, our brains are still wired to focus
[00:18:02] on potential threats. The brain detects dangers more
[00:18:06] quickly, processes them more intensely, and stores them more
[00:18:09] deeply. The predators might be gone, but in our minds they're
[00:18:13] still alive and well.
[00:18:16] Negative things stick more strongly. Studies show that
[00:18:20] even individual words like crisis or danger are
[00:18:23] enough to make us click a headline. Our eyes are
[00:18:27] drawn to them, and before we know it, we've gone down another
[00:18:30] rabbit hole.
[00:18:34] Large scale data analysis also shows that online head.
[00:19:07] So the media amplifies our perception. Does that mean the
[00:19:11] more news we consume, the worse the world
[00:19:13] seems? In today's world, algorithms play
[00:19:17] a role.
[00:19:23] determined by what friends have been up to, but also by what the
[00:19:27] algorithm adds in, and it's not
[00:19:29] accidental.
[00:19:32] Negative information has a longer dwell time and
[00:19:35] leads to more interaction. Algorithms
[00:19:43] fuel what our brains prefer, the dramatic. But if we only
[00:19:47] focus on crises, we lose our inner
[00:19:49] compass, and our view of the world becomes
[00:19:53] distorted.
[00:19:56] It's a vicious circle in which we seem to live in a world
[00:19:59] dominated by crisis.
[00:20:02] But it's not that there are fewer positive developments out there,
[00:20:06] they're just less visible.
[00:20:50] is too. A persistent negative focus fuels
[00:20:53] mistrust, cynicism, and disillusionment with
[00:20:56] politics. Our world view is no longer shaped just by
[00:21:00] facts, but also by filters, emotions, and feedback
[00:21:03] loops. So the crucial question isn't how bad is the
[00:21:07] world, but rather how can we learn to see things more
[00:21:10] realistically?
[00:21:17] in the digital age,
[00:21:50] and to life itself. And being critical doesn't have to be negative, it
[00:21:54] can also mean that I've assessed it, and it's not so bad after
[00:21:57] all.
[00:22:00] The world is certainly not perfect, but it's far from all
[00:22:03] bad either. Sometimes it just needs a new
[00:22:07] headline.
[00:22:12] Lead poisoning used to be widespread in industrialized
[00:22:15] countries, but that's changed. The drastic reduction in environmental
[00:22:20] exposure is regarded as one of the biggest public health achievements of the 20th
[00:22:24] century.
[00:22:29] In the early 1970s, in Germany, lead was as omnipresent as cigarette
[00:22:33] smoke in pubs, found in everything from paints and drinking water to
[00:22:37] gasoline, and most people didn't see it as a
[00:22:40] problem. When I was a boy, casting melted lead
[00:22:43] shapes was a normal game on New Year's Eve. For
[00:22:47] centuries, people treasured lead's all round.
[00:23:20] to its effects on the brain in a moment, but
[00:23:23] first, how does it enter our bodies in the first
[00:23:26] place? Via food, the
[00:23:29] air, or drinking water. And once lead is in the
[00:23:33] body, it doesn't leave quickly. Lead has a surprisingly long half
[00:23:36] life in bones, up to 30 years, so even three decades
[00:23:40] after intake, about half of it is still
[00:23:43] there. And the more lead that accumulates in the
[00:23:46] body, the greater the damage. With an
[00:23:49] increased exposure affects various organ
[00:23:52] systems, the nervous system, and also blood cell
[00:23:56] production, the kidney, and the cardiovascular
[00:23:59] system. The International Agency for Research on
[00:24:03] Cancer even classifies lead as probably
[00:24:06] carcinogenic. And children are especially vulnerable. While playing
[00:24:10] in sandboxes, they'll often put their hands in their mouths, together with any
[00:24:14] lead residue in the sand, and their bodies
[00:24:17] absorb the metal more efficiently. The greatest.
[00:24:50] In West Germany, as in some other countries, the government responded by
[00:24:53] gradually reducing the lead content in gasoline. By
[00:24:56] 1988, the permitted lead content of regular
[00:24:59] gasoline was zero,
[00:25:02] followed in 1996 by premium gasoline.
[00:25:06] Since 2021, leaded gasoline has been banned
[00:25:09] worldwide. By 2024,
[00:25:12] 48% of countries had introduced regulations on
[00:25:15] lead-based paints, and in Germany all lead
[00:25:19] pipes had to be decommissioned or removed by mid-January
[00:25:22] 2026. How effective those measures were can
[00:25:26] be seen in lead concentrations of blood samples. A
[00:25:29] comparison between values from the 1980s and present day
[00:25:33] data shows a clear
[00:25:35] trend.
[00:25:38] Over the past decades, the ban on leaded gasoline has resulted in a
[00:25:41] significant drop in blood lead levels. We have
[00:25:45] data from students in Muenster, for example, who are examined in groups
[00:25:49] by the.
[00:25:59] down to 10 micrograms per liter by 2018.
[00:30:27] Good evening, this is another Red Line, I'm Marina Fratucan. Around 20
[00:30:31] women are killed annually in each of the countries in the region, and another about
[00:30:34] 15 of them were exposed to attempted murder. In only
[00:30:37] 6% of cases, the perpetrator is unknown. However, murders are not
[00:30:41] prevented. Is femicide predictable? Is there a typical
[00:30:45] profile of an abuser-killer? Does the system recognize
[00:30:48] the risk to the victim, or is femicide a systemic failure?
[00:30:51] Will femicide become a separate criminal offense in Serbia and Republika
[00:30:55] Srpska, as it has been since 2024 in Croatia
[00:30:58] and last year in the Federation?